阅读说明：本技术 诊断辅助程序 (Diagnostic aid ) 是由 阿部武彦 吉田典史 于 2019-01-07 设计创作，主要内容包括：本发明提供了一种能显示针对包含呼气或吸气的全部或一部分的每个呼吸要素而形状发生变化的区域的运动的诊断辅助程序。包含下述处理：从存储图像的数据库中获取多帧图像；基于各帧图像的特定区域的像素来确定包含呼气或吸气的全部或一部分的呼气要素的周期；基于所确定出的呼吸要素的周期,来检测肺野；将检测出的肺野分割成多个块区域,计算各帧图像中的块区域的图像的变化；对各帧图像中的各块区域的图像的变化进行傅里叶变换；从傅里叶变换后所得到的频谱中提取包含与呼吸要素的周期相对应的频谱在内的固定频带内的频谱；对从所述固定频带中提取出的频谱进行反傅里叶变换；以及将反傅里叶变换后的各图像显示在显示器上。(The present invention provides a diagnostic support program capable of displaying the movement of a region of which the shape changes for each respiratory element including all or a part of expiration or inspiration. Comprises the following steps: acquiring a plurality of frame images from a database for storing the images; determining a period of an expiratory element including all or a part of expiration or inspiration based on pixels of a specific region of each frame image; detecting a lung field based on the determined period of the respiratory element; dividing the detected lung fields into a plurality of block regions, and calculating the change of the image of the block region in each frame image; fourier transform is carried out on the change of the image of each block area in each frame image; extracting a spectrum in a fixed frequency band including a spectrum corresponding to a period of a respiratory element from a spectrum obtained after Fourier transform; performing inverse fourier transform on a frequency spectrum extracted from the fixed frequency band; and displaying each image after the inverse Fourier transform on a display.)

1. A diagnostic aid program for assisting in the diagnosis of a patient,

the diagnosis assisting program analyzes an image of a human body and displays an analysis result, and is characterized by causing a computer to execute:

acquiring a plurality of frames of images from a database for storing the images;

a process of determining a frequency of at least one respiratory element among respiratory elements including all or part of exhalation or inhalation based on pixels of a specific region of each frame image;

a process of detecting a lung field based on a frequency of at least one of the determined respiratory elements;

a process of dividing the detected lung field into a plurality of block regions and calculating a change in image of the block regions in each of the frame images;

Processing fourier transform of changes in the image of each block region in each frame image;

a process of extracting a spectrum in a fixed frequency band including a spectrum corresponding to a frequency of at least one of the respiratory elements from the spectrum obtained after the fourier transform;

performing an inverse fourier transform process on a spectrum extracted from the fixed frequency band; and

and displaying each image after the inverse Fourier transform on a display.

2. The diagnostic assistance program of claim 1,

further comprising the following processes: and a process of extracting, using a filter, a spectrum in a fixed band including a frequency of noise and a spectrum in a fixed band including a spectrum corresponding to a frequency other than a frequency of a respiratory component, or an input frequency or band obtained from the frame image, from the spectrum obtained after the fourier transform.

3. The diagnostic assistance program of claim 1 or 2,

further comprising the following processes: and a process of generating an image between the frames based on the frequency of the respiratory component and the images of the frames.

4. A diagnostic aid program for assisting in the diagnosis of a patient,

the diagnosis assisting program analyzes an image of a human body and displays an analysis result, and is characterized by causing a computer to execute:

acquiring a plurality of frames of images from a database for storing the images;

a process of determining a frequency of at least one of cardiovascular pulsation components extracted from cardiac beats or vascular pulsations of a subject;

a process of determining a frequency of at least one respiratory element among respiratory elements including all or part of exhalation or inhalation based on pixels of a specific region of each frame image;

a process of detecting a lung field based on a frequency of at least one of the determined respiratory elements;

a process of dividing the detected lung field into a plurality of block regions and calculating a change in image of the block regions in each of the frame images;

processing fourier transform of changes in the image of each block region in each frame image;

extracting a spectrum in a fixed frequency band including a spectrum corresponding to a frequency of at least one of the cardiovascular beat elements from the spectrum obtained after the fourier transform;

Performing an inverse fourier transform process on a spectrum extracted from the fixed frequency band; and

and displaying each image after the inverse Fourier transform on a display.

5. A diagnostic aid program for assisting in the diagnosis of a patient,

the diagnosis assisting program analyzes an image of a human body and displays an analysis result, and is characterized by causing a computer to execute:

acquiring a plurality of frames of images from a database for storing the images;

a process of determining a frequency of at least one of cardiovascular pulsation components extracted from cardiac beats or vascular pulsations of a subject;

processing for detecting lung fields;

a process of dividing the detected lung field into a plurality of block regions and calculating a change in image of the block regions in each of the frame images;

processing fourier transform of changes in the image of each block region in each frame image;

extracting a spectrum in a fixed frequency band including a spectrum corresponding to a frequency of at least one of the cardiovascular beat elements from the spectrum obtained after the fourier transform;

performing an inverse fourier transform process on a spectrum extracted from the fixed frequency band; and

And displaying each image after the inverse Fourier transform on a display.

6. The diagnostic assistance program of claim 4 or 5,

further comprising the following processes: and a process of extracting, using a filter, a spectrum in a fixed band including a frequency of noise and a spectrum in a fixed band including a spectrum corresponding to a frequency other than the frequency of the cardiovascular pulse component obtained from the frame image, or an input frequency or band, from the spectrum obtained after the fourier transform.

7. The diagnostic assistance procedure of any one of claims 4 to 6,

further comprising the following processes: and generating the images between the frames based on the determined frequency of the cardiovascular pulsation component and the images of the frames.

8. A diagnostic aid program for assisting in the diagnosis of a patient,

the diagnosis assisting program analyzes an image of a human body and displays an analysis result, and is characterized by causing a computer to execute:

acquiring a plurality of frames of images from a database for storing the images;

a process of determining a frequency of at least one blood vessel pulsation element among blood vessel pulsation elements extracted from blood vessel pulsation of a subject;

A process of dividing an analysis range set for each of the frame images into a plurality of block regions and calculating a change in image of the block region in each of the frame images;

processing fourier transform of changes in the image of each block region in each frame image;

extracting a spectrum in a fixed frequency band including a spectrum corresponding to a frequency of at least one of the blood vessel pulsation elements from the spectrum obtained after the fourier transform;

performing an inverse fourier transform process on a spectrum extracted from the fixed frequency band; and

and displaying each image after the inverse Fourier transform on a display.

9. The diagnostic assistance program of claim 8,

further comprising the following processes: and a process of extracting, using a filter, a spectrum in a fixed band including a frequency of noise and a spectrum in a fixed band including a spectrum corresponding to a frequency other than a frequency of a blood vessel pulsation component, or an input frequency or band obtained from the frame image, from the spectrum obtained after the fourier transform.

10. The diagnostic assistance program of claim 8 or 9,

further comprising the following processes: and generating the images between the frames based on the determined frequency of the cardiovascular pulsation component and the images of the frames.

11. A diagnostic aid program for assisting in the diagnosis of a patient,

the diagnosis assisting program analyzes an image of a human body and displays an analysis result, and is characterized by causing a computer to execute:

acquiring a plurality of frames of images from a database for storing the images;

a process of determining a frequency of at least one respiratory element among respiratory elements including all or part of exhalation or inhalation based on pixels of a specific region of each frame image;

a process of detecting a lung field and a diaphragm based on a frequency of at least one of the determined respiratory elements;

a process of dividing the detected lung field into a plurality of block regions and calculating a change rate of pixels of the block regions in each of the frame images;

a process of extracting only a block region having an adjustable rate within a predetermined fixed range using an adjustable rate that is a value of a ratio of a rate of change of pixels of the block region to a rate of change of a moving portion in conjunction with respiration; and

Displaying each image including only the extracted block region on a display.

12. The diagnostic assistance program of claim 11,

further comprising the following processes: and a process for determining the frequency of at least one of the cardiovascular pulsation elements extracted from the heartbeat or the blood vessel pulsation of the subject, or the frequency of at least one of the blood vessel pulsation elements extracted from the blood vessel pulsation.

13. The diagnostic assistance program of claim 11 or 12,

the value of the logarithm of the adjustable rate is determined as a fixed range including 0.

14. The diagnostic assistance procedure of any one of claims 1 to 13,

further comprising the following processes: the lung fields in other frames are detected using at least one Bezier curve (Bezier curve) on the lung fields detected in a specific frame.

15. The diagnostic assistance program of claim 14,

an internal control point is selected within the detected lung fields, and the lung fields are segmented using a curve or a straight line passing through the internal control point within the lung fields.

16. The diagnostic assistance program of claim 15,

the intervals between the control points on the extension of the detected lung fields and the vicinity thereof are made relatively large, and the intervals between the internal control points are made relatively small in accordance with the expansion ratio of each portion in the detected lung fields.

17. The diagnostic assistance program of claim 15,

in the detected lung fields, the intervals of the control points are made relatively large as they proceed in the head-tail direction with respect to the human body, or the intervals of the control points are made relatively large as they proceed in a specific vector direction.

18. The diagnostic assistance procedure of any one of claims 1 to 13,

further comprising the following processes: the lung fields in other frames are detected using at least one Bezier curve (Bezier curve) on the lung fields detected in a specific frame.

19. The diagnostic assistance procedure of any one of claims 1 to 13,

further comprising the following processes: at least one Bezier curve (Bezier curve) is used to detect the range corresponding to the analysis range in other frames in the analysis range predetermined in a specific frame.

20. The diagnostic assistance procedure of any one of claims 1 to 13,

further comprising the following processes: at least one Bezier curve (Bezier curve) is used to perform a process of rendering at least the lung field, the blood vessel or the heart.

21. A diagnostic aid program for assisting in the diagnosis of a patient,

the diagnosis assisting program analyzes an image of a human body and displays an analysis result, and is characterized by causing a computer to execute:

acquiring a plurality of frames of images from a database for storing the images;

processing for determining an analysis range using a bezier curve for all the acquired frame images; and

processing of detecting an analysis object based on a change in intensity (intensity) within the analysis range.

22. The diagnostic assistance program of claim 21,

further comprising a process of calculating a feature of the detected edge of the analysis object.

23. The diagnostic assistance procedure of any one of claims 1 to 22,

for each of the successive images, the diaphragm is detected by calculating the difference in intensity (intensity), and

and displaying an index indicating the position or shape of the detected diaphragm or the moving part linked with respiration.

24. The diagnostic assistance program of claim 23,

by changing the threshold value of the intensity (intensity), the diaphragm that is not blocked by the portion other than the diaphragm is displayed, and the entire shape of the diaphragm is interpolated.

25. The diagnostic assistance program of claim 23 or 24,

further comprising the following processes: and a process of calculating a frequency of at least one of the respiratory elements from the detected position or shape of the diaphragm or the position or shape of the moving part linked with respiration.

26. The diagnostic assistance procedure of any one of claims 1 to 22,

further comprising the following processes: and (3) a process of spatially normalizing the detected lung fields, or temporally normalizing the detected lung fields by reconstruction (reconstruction).

27. The diagnostic assistance procedure of any one of claims 1 to 22,

the respiratory element is corrected by changing the phase of the frequency of at least one of the respiratory elements or smoothing the waveform of the respiratory element.

28. The diagnostic assistance procedure of any one of claims 1 to 27,

a waveform of any portion within an analysis range is specified, a component of the frequency of the specified waveform is extracted, and an image corresponding to the component of the frequency of the waveform is output.

29. The diagnostic assistance procedure of any one of claims 1 to 28,

the density (density) of the analytical range is detected and the locations where relatively large changes in density occur are removed.

30. The diagnostic assistance procedure of any one of claims 1 to 29,

further comprising the following processes: and a processing of selecting at least one frequency when performing inverse fourier transform based on a spectral composition ratio of a periodic change specific to an organ from among the frequency spectra obtained after the fourier transform.

31. The diagnostic assistance procedure of any one of claims 1 to 30,

and controlling an X-ray imaging device according to the frequency of at least one respiratory element in the respiratory elements so as to adjust the irradiation interval of the X-rays.

32. The diagnostic assistance procedure of any one of claims 1 to 10,

After the inverse fourier transform, only a block having a relatively large amplitude value is extracted and displayed.

33. The diagnostic assistance procedure of any one of claims 1 to 32,

further comprising the following processes: and after the lung fields are identified, determining a diaphragm or a thorax, calculating the variation of the diaphragm or the thorax, and calculating the change rate according to the variation.

34. The diagnostic assistance procedure of any one of claims 1 to 32,

the method further includes a process of multiplying a coefficient by a specific spectrum, and performs highlight display based on the specific spectrum obtained by multiplying the coefficient.

35. The diagnostic assistance procedure of any one of claims 1 to 34,

after acquiring a plurality of frame images from a database storing images, digital filtering is performed on a site to be analyzed in order to determine the frequency or waveform of a respiratory component.

36. The diagnostic assistance procedure of any one of claims 1 to 35,

determining a plurality of frequencies of respiratory elements including all or a part of exhalation or inhalation based on pixels of a specific region of each frame image, and

Displaying, on a display, respective images corresponding to respective ones of a plurality of frequencies of the respiratory element.

37. The diagnostic assistance procedure of any one of claims 1 to 35,

for a specific range of one or more frame images, images aggregated to a fixed value are selected and displayed on a display.

Technical Field

The present invention relates to a technique for analyzing an image of a human body and displaying the analysis result.

Background

When a doctor diagnoses a lung by using a dynamic chest image, it is important to observe a time-series dynamic chest image obtained by imaging an object in a natural breathing state. Methods for evaluating lung function, such as spirometers, which easily obtain physiological data, RI (Radio Isotope) examinations, conventional X-ray photographs, which can obtain morphological data, and CT (Computed Tomography), are known. However, it is not easy to efficiently acquire both physiological data and morphological data.

In recent years, a method of taking a moving image of a human chest using a semiconductor image sensor such as an FPD (Flat panel detector) for diagnosis has been attempted. For example, non-patent document 1 discloses a technique of: a difference image representing the difference between signal values is generated between a plurality of frame images constituting a moving image, and the maximum value of each signal value is obtained from the difference image and displayed.

Further, patent document 1 discloses a technique of: a lung field region is extracted from each frame of images of a plurality of frames representing the dynamic state of the chest of a human body, the lung field region is divided into a plurality of small regions, and the divided small regions are correlated with each other between the plurality of frames of images and analyzed. According to this technique, a feature quantity indicating the motion of a small region obtained by segmentation is displayed.

Disclosure of Invention

Technical problem to be solved by the invention

However, as in the technique described in non-patent document 1, it is not easy for a doctor to grasp a medical condition simply by displaying the maximum value of the inter-frame difference value of each pixel of a moving image. Further, as in the technique described in patent document 1, it is not sufficient to grasp the state of illness only by displaying the feature amount. Therefore, it is desirable to display an image corresponding to the state of the respiratory and pulmonary vessels. That is, it is desirable to grasp the breathing state and the whole blood vessel dynamic state of the subject, that is, the human body, and display an image representing the actual motion based on the waveform or frequency of the respiration, the blood vessel or blood flow in the heart or the pulmonary gate part, or the tendency of the image to change.

The present invention has been made in view of the above circumstances, and an object thereof is to provide a diagnostic support program capable of displaying the movement of a region having a shape that changes for each respiratory element including all or a part of exhalation or inhalation. More specifically, the object is to calculate a numerical value as a diagnosis aid by quantifying the coincidence of a waveform and Hz or other inconsistencies with respect to a new object to be measured, and to generate an image as a diagnosis aid by further imaging the numerical value.

Technical scheme for solving technical problem

(1) In order to achieve the above object, the present invention adopts the following technical means. That is, a diagnosis support program according to an aspect of the present invention is a diagnosis support program for analyzing an image of a human body and displaying an analysis result, the diagnosis support program causing a computer to execute: acquiring a plurality of frames of images from a database for storing the images; a process of determining a frequency of at least one of expiratory elements including all or a part of expiration or inspiration based on pixels of a specific region of each frame image; a process of detecting a lung field based on a frequency of at least one expiratory element of the determined respiratory elements; a process of dividing the detected lung field into a plurality of block regions and calculating a change in image of the block regions in each of the frame images; processing fourier transform of changes in the image of each block region in each frame image; a process of extracting a spectrum in a fixed frequency band including a spectrum corresponding to a frequency of at least one of the respiratory elements from the spectrum obtained after the fourier transform; performing an inverse fourier transform process on a spectrum extracted from the fixed frequency band; and a process of displaying each image after the inverse fourier transform on a display.

(2) In addition, a diagnosis support program according to an aspect of the present invention further includes: and a process of extracting, using a filter, a spectrum in a fixed band including a frequency of noise and a spectrum in a fixed band including a spectrum corresponding to a frequency other than a frequency of a respiratory component obtained from the frame image or an input frequency or band from the spectrum obtained after the fourier transform.

(3) In addition, a diagnosis support program according to an aspect of the present invention further includes: and a process of generating an image between the frames based on the frequency of the respiratory component and the images of the frames.

(4) A diagnosis support program according to an aspect of the present invention is a diagnosis support program for analyzing an image of a human body and displaying an analysis result, the diagnosis support program causing a computer to execute: acquiring a plurality of frames of images from a database for storing the images; a process of determining a frequency of at least one blood vessel pulsation element among the cardiovascular pulsation elements extracted from the heart beat or the blood vessel pulsation of the subject; a process of determining a frequency of at least one of expiratory elements including all or a part of expiration or inspiration based on pixels of a specific region of each frame image; a process of detecting a lung field based on a frequency of at least one expiratory element of the determined respiratory elements; a process of dividing the detected lung field into a plurality of block regions and calculating a change in image of the block regions in each of the frame images; processing fourier transform of changes in the image of each block region in each frame image; extracting a spectrum in a fixed frequency band including a spectrum corresponding to a frequency of at least one of the cardiovascular beat elements from the spectrum obtained after the fourier transform; performing an inverse fourier transform process on a spectrum extracted from the fixed frequency band; and a process of displaying each image after the inverse fourier transform on a display.

(5) A diagnosis support program according to an aspect of the present invention is a diagnosis support program for analyzing an image of a human body and displaying an analysis result, the diagnosis support program causing a computer to execute: acquiring a plurality of frames of images from a database for storing the images; a process of determining a frequency of at least one of cardiovascular pulsation components extracted from cardiac beats or vascular pulsations of a subject; processing for detecting lung fields; a process of dividing the detected lung field into a plurality of block regions and calculating a change in image of the block regions in each of the frame images; processing fourier transform of changes in the image of each block region in each frame image; extracting a spectrum in a fixed frequency band including a spectrum corresponding to a frequency of at least one of the cardiovascular beat elements from the spectrum obtained after the fourier transform; performing an inverse fourier transform process on a spectrum extracted from the fixed frequency band; and a process of displaying each image after the inverse fourier transform on a display.

(6) In addition, a diagnosis support program according to an aspect of the present invention further includes: and a process of extracting, using a filter, a spectrum in a fixed band including a frequency of noise and a spectrum in a fixed band including a spectrum corresponding to a frequency other than a frequency of the cardiovascular pulsation component obtained from the frame image or an inputted frequency or band from the spectrum obtained after the fourier transform.

(7) In addition, a diagnosis support program according to an aspect of the present invention further includes: and generating the images between the frames based on the determined frequency of the cardiovascular pulsation component and the images of the frames.

(8) A diagnosis support program according to an aspect of the present invention is a diagnosis support program for analyzing an image of a human body and displaying an analysis result, the diagnosis support program causing a computer to execute: acquiring a plurality of frames of images from a database for storing the images; a process of determining a frequency of at least one of cardiovascular pulsation components extracted from cardiac beats or vascular pulsations of a subject; a process of dividing an analysis range set for each of the frame images into a plurality of block regions and calculating a change in image of the block region in each of the frame images; processing fourier transform of changes in the image of each block region in each frame image; extracting a spectrum in a fixed frequency band including a spectrum corresponding to a frequency of at least one of the cardiovascular beat elements from the spectrum obtained after the fourier transform; performing an inverse fourier transform process on a spectrum extracted from the fixed frequency band; and a process of displaying each image after the inverse fourier transform on a display.

(9) In addition, a diagnosis support program according to an aspect of the present invention further includes: and a process of extracting, using a filter, a spectrum in a fixed band including a frequency of noise and a spectrum in a fixed band including a spectrum corresponding to a frequency other than a frequency of the cardiovascular pulsation component obtained from the frame image or an inputted frequency or band from the spectrum obtained after the fourier transform.

(10) In addition, a diagnosis support program according to an aspect of the present invention further includes: and generating the images between the frames based on the determined frequency of the cardiovascular pulsation component and the images of the frames.

(11) A diagnosis support program according to an aspect of the present invention is a diagnosis support program for analyzing an image of a human body and displaying an analysis result, the diagnosis support program causing a computer to execute: acquiring a plurality of frames of images from a database for storing the images; a process of determining a frequency of at least one of expiratory elements including all or a part of expiration or inspiration based on pixels of a specific region of each frame image; a process of detecting a lung field and a diaphragm based on a frequency of at least one of the determined respiratory elements; a process of dividing the detected lung field into a plurality of block regions and calculating a change rate of pixels of the block regions in each of the frame images; extracting only a block region having an adjustable rate within a predetermined fixed range using an adjustable rate that is a value of a ratio of a rate of change of pixels of the block region image to a rate of change of a moving portion in conjunction with respiration; and a process of displaying each image including only the extracted block region on a display.

(12) In addition, a diagnosis support program according to an aspect of the present invention further includes: and a process for determining the frequency of at least one of the cardiovascular pulsation elements extracted from the cardiac beat or the blood vessel pulsation of the subject or the frequency of at least one of the blood vessel pulsation elements extracted from the blood vessel pulsation.

(13) In the diagnosis support program according to one aspect of the present invention, the logarithm of the adjustable rate is determined to be a fixed range including 0.

(14) In addition, a diagnosis support program according to an aspect of the present invention further includes: the lung fields in other frames are detected using at least one bezier curve (bezier curve) on the lung fields detected in a specific frame.

(15) In the diagnosis support program according to one aspect of the present invention, an internal control point is selected in the detected lung fields, and the lung fields are divided by a curve or a straight line passing through the internal control point in the lung fields.

(16) In the diagnosis support program according to one aspect of the present invention, the intervals between the control points on the extension of the detected lung fields and in the vicinity thereof are made relatively large, and the intervals between the internal control points are made relatively small in accordance with the expansion ratio of each part in the detected lung fields.

(17) In the diagnosis support program according to one aspect of the present invention, the intervals of the control points in the detected lung fields are made relatively large as they move forward from the head-to-tail direction with respect to the human body, or the intervals of the control points are made relatively large as they move along a specific vector direction.

(18) In addition, a diagnosis support program according to an aspect of the present invention further includes: the lung fields in other frames are detected using at least one bezier curve (bezier curve) on the lung fields detected in a specific frame.

(19) In addition, a diagnosis support program according to an aspect of the present invention further includes: at least one Bezier curve (Bezier curve) is used in an analysis range predetermined in a specific frame to detect a range corresponding to the analysis range in another frame.

(20) In addition, a diagnosis support program according to an aspect of the present invention further includes: at least one Bezier curve (Bezier curve) is used to perform a process of rendering at least the lung field, the blood vessel or the heart.

(21) A diagnosis support program according to an aspect of the present invention is a diagnosis support program for analyzing an image of a human body and displaying an analysis result, the diagnosis support program causing a computer to execute: acquiring a plurality of frames of images from a database for storing the images; processing for determining an analysis range using a bezier curve for all the acquired frame images; and a process of detecting an analysis target based on a change in intensity (intensity) within the analysis range.

(22) In addition, a diagnosis support program according to an aspect of the present invention further includes a process of calculating a feature of the detected edge of the analysis target.

(23) In addition, a diagnosis support program according to an aspect of the present invention is characterized in that a diaphragm is detected by calculating a difference in intensity (intensity) for each of continuous images, and an index indicating a position or a shape of the detected diaphragm or a moving portion linked with respiration is displayed.

(24) In addition, a diagnosis support program according to an aspect of the present invention is characterized in that the threshold value of the intensity (intensity) is changed to display a diaphragm that is not blocked by a portion other than the diaphragm, and the entire shape of the diaphragm is interpolated.

(25) In addition, a diagnosis support program according to an aspect of the present invention further includes: and a process of calculating a frequency of at least one of the respiratory elements from the detected position or shape of the diaphragm or the position or shape of the moving part linked with respiration.

(26) In addition, a diagnosis support program according to an aspect of the present invention further includes: and (3) a process of spatially normalizing the detected lung fields, or temporally normalizing the detected lung fields by reconstruction (reconstruction).

(27) In the diagnosis support program according to one aspect of the present invention, the respiratory component is corrected by changing a phase of a frequency of at least one of the respiratory components or smoothing a waveform of the respiratory component.

(28) In addition, a diagnosis support program according to an aspect of the present invention specifies a waveform of any portion within an analysis range, extracts a component of a frequency of the specified waveform, and outputs an image corresponding to the component of the frequency of the waveform.

(29) In addition, a diagnosis support program according to an aspect of the present invention is characterized in that the density (density) of the analysis range is detected and a position where a relatively large change in density occurs is removed.

(30) In addition, a diagnosis support program according to an aspect of the present invention further includes: and a processing of selecting at least one frequency when performing inverse fourier transform based on a spectral composition ratio of a periodic change specific to an organ from among the frequency spectra obtained after the fourier transform.

(31) In addition, a diagnosis support program according to an aspect of the present invention is characterized in that the X-ray imaging device is controlled to adjust an irradiation interval of the X-rays in accordance with a frequency of at least one of the respiratory elements.

(32) In the diagnosis support program according to one aspect of the present invention, after the inverse fourier transform, only blocks having relatively large amplitude values are extracted and displayed.

(33) In addition, a diagnosis support program according to an aspect of the present invention further includes: after the lung fields are identified, a diaphragm or a thorax is determined, a variation amount of the diaphragm or the thorax is calculated, and a change rate is calculated according to the variation amount.

(34) In addition, a diagnosis support program according to an aspect of the present invention further includes a process of multiplying a coefficient by a specific spectrum, and performs an emphasized display based on the specific spectrum obtained by multiplying the coefficient.

(35) In addition, a diagnosis support program according to an aspect of the present invention is characterized in that after acquiring a plurality of frame images from a database storing images, digital filtering is applied to a region to be analyzed in order to specify a frequency or a waveform of a respiratory component.

(36) In addition, a diagnosis support program according to an aspect of the present invention further includes: a plurality of frequencies of a respiratory element including all or a part of exhalation or inhalation are determined based on pixels of a specific region of each frame image, and each image corresponding to each of the plurality of frequencies of the respiratory element is displayed on a display.

(37) In addition, a diagnosis support program according to an aspect of the present invention is characterized in that images grouped to a certain fixed value are selected for a specific range of one or more frame images and displayed on a display.

Effects of the invention

According to one embodiment of the present invention, the movement of the region having the changed shape for each respiratory element including all or part of exhalation or inhalation can be displayed.

Drawings

Fig. 1A is a diagram showing a schematic configuration of a diagnosis support system according to the present embodiment.

Fig. 1B is a diagram showing an example of a method of segmenting a lung region.

Fig. 1C is a diagram showing changes in lung morphology with time.

Fig. 1D is a diagram showing changes in lung morphology with time.

Fig. 2A is a graph showing the "intensity" change of a specific block and the result obtained by fourier analysis thereof.

Fig. 2B is a diagram showing fourier transform results obtained by extracting frequency components close to the heart beat and changes in "intensity" of the frequency components close to the heart beat obtained by performing inverse fourier transform on the fourier transform results.

Fig. 2C is a diagram showing an example of extracting a certain fixed band from a spectrum obtained after fourier transform.

Fig. 2D is a graph schematically showing the rate of change of the lung.

Fig. 2E is a diagram showing an example of a pattern image of the lung field region.

Fig. 2F is a diagram showing an example of a pattern image of the lung field region.

Fig. 2G is a diagram showing an example of a pattern image of the lung field region.

Fig. 2H is a diagram showing an example of a pattern image of the lung field region.

Fig. 3A is a diagram showing an example in which the contour of the lung field is drawn using both bezier curves and straight lines, and shows a state in which the lung field is maximized.

Fig. 3B is a diagram showing an example in which the contour of the lung field is drawn using both bezier curves and straight lines, and shows a state in which the lung field is minimal.

Fig. 4A is a diagram in which the front and rear of the images of the lung fields are superimposed between the previous frame and the next frame.

Fig. 4B is a diagram showing a state in which the differences between the 2 original images of fig. 4A are acquired, and as a result, "lines with strong gaps" are generated.

Fig. 4C is a diagram showing a difference value of the total "density" of the "intensity" values at each position in the vertical direction of the image in fig. 4B.

Fig. 5 is a graph showing the result of curve regression and approximation of the relative position of the diaphragm.

Fig. 6A is a flowchart showing an outline of the respiratory function analysis according to the present embodiment.

Fig. 6B is a diagram showing an example of an image displayed on the display.

Fig. 6C is a diagram showing an example of an image displayed on the display.

Fig. 7 is a flowchart showing an outline of the pulmonary blood flow analysis according to the present embodiment.

Fig. 8 is a flowchart showing an outline of another blood flow analysis according to the present embodiment.

Fig. 9 is a diagram showing an example in which a coefficient is multiplied by a fixed spectrum in a spectrum obtained after fourier transform.

Fig. 10 is a diagram showing an example of plotting the lung fields using bezier curves.

Fig. 11 is a diagram showing an example of dividing lung fields using bezier curves.

Fig. 12 is a diagram showing an example in which lung fields are divided using bezier curves.

Fig. 13 is a diagram showing an example of comparing a waveform of an aortic blood flow volume and a waveform of a ventricular volume.

Fig. 14 is a diagram showing an example of a lung and a pixel value in the vicinity of the lung.

Fig. 15 is a diagram schematically showing a schematic structure of a blood vessel of a human body.

Detailed Description

First, the basic concept of the present invention will be explained. In the present invention, a motion repeatedly captured at a fixed cycle among respiration in a human body, the area and volume of a blood vessel, a lung field, and other biological motions is captured and measured as a wave in whole or in a part of the range, the motion being fixed or repeated on the time axis (a set of motions). For the measurement results of the waves, the morphology of the (first) wave itself or the spacing of the (second) wave (frequency: Hz) is used. These 2 concepts are collectively referred to as "base data".

There may be waves at the same time that are linked in the same way. For example, for the case of breathing, the following approximation may be considered.

(average of "density" variation over a certain approximate range) about (variation of thoracic cage) about (motion of diaphragm) about (lung function inspection) about (about channel of the back and forth respiration sensor)

For the above-mentioned "(form of the first Wave itself)," waveform tunability "is used as a concept, and an image is displayed based on this (Wave form tunable imaging). Further, the concept of "Frequency tunability" is used for the "interval of two waves (Frequency: Hz)", and an image is displayed based on this (Frequency tunable imaging).

For example, in the case of the heart, as shown in fig. 13, which is an "example in which the waveform of the aortic blood flow and the waveform of the ventricular volume are compared", the peak of the aortic blood flow does not coincide with the peak of the ventricular volume and the waveform. However, in fig. 13, if the time widths of equal intervals of time t1 to t2, time t2 to t3, and time t3 to t4 … are determined to be 1 cycle, 1 cycle of the aortic blood flow and 1 cycle of the ventricular volume are repeated a plurality of times, and it can be said that the frequency of each waveform is adjustable. Focusing on this waveform, a waveform (Wave form) can be expected by determining 1 cycle from actual measurement values as shown in fig. 13 and using a model waveform. That is, as a generation method of the "waveform as the basic data", actual measurement may be performed, generation may be performed according to a frequency (cycle), a model waveform may be used, or an inter-individual waveform may be averaged and used. Since a waveform (Wave form) can be expected if the cycle (cycle) of an organ having a frequency such as a heart is known, it is possible to grasp the waveform of an aortic blood flow, a ventricular volume, and the like, and display an image of the movement of the organ based on the waveform.

In acquiring the change in "density" of the breath, heart, lung, and the like, a digital filter may be applied in advance so that other elements are not mixed.

In the present invention, the concept of "respiratory element" is used. The term "respiratory element" means all or a part of exhalation or inhalation. For example, it is considered that "1 breath" is divided into "1 exhalation" and "1 inhalation", and "1 breath" may be limited to any one of "0%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100%" of "1 exhalation or 1 inhalation". Further, only a fixed proportion of each expired air, for example, only 10% of the expired air may be extracted and evaluated. By using data of any one of them or data obtained by combining them, extraction of an image with higher accuracy is realized. At this time, the calculations may be performed a plurality of times with each other.

Such a consideration can be applied not only to "respiratory elements" but also to "cardiovascular elements" as well.

Here, when generating the basic data, the accuracy is improved by supplementing the extraction of components with each other by a plurality of waveform measurements of characteristic quantities (for example, 2 or more kinds of the "density" and the "volume measurement" in a certain fixed range, the motion of the thorax, the motion of the diaphragm, the "spirometry", and the thoracoabdominal respiration sensor) obtained from a single or a plurality of modalities, or the same respiratory cycle. This can reduce artifacts and improve accuracy based on a certain fixed assumption such as a line (line). Here, the "density (density)" means, in an image, an "absorption value" of a pixel in a specific region, although it can be interpreted as "density (mitsudo)". For example, in CT, air is used as "-1000", bone as "1000" and water as "0".

Further, the axis, width, range, and fluctuation in Hz and width of the wave obtained by the component extraction from each other are estimated. That is, the axes of Hz are set to be averaged by a plurality of times of superposition, and the optimal intervals (ranges) of axes, widths, ranges, and Hz are calculated by variance. At this time, there are cases where: hz (noise) of other actions is extracted, and if the waves exist, the Hz is relatively measured even to the extent that the waves are not included. That is, only a part of the waveform may be extracted from the entire waveform element.

In the present specification, "density" and "strength" are used differently. As described above, "intensity" means an absorption value, and when a numerical value is given to a case where the air permeability is high in the original image of the XP or XP video and the portion with high permeability is white, the air is displayed as "-1000", the water is displayed as "0", and the bone is displayed as "1000". On the other hand, "intensity" is changed with respect to "density", and is, for example, normalized to "convert" into a width of concentration, a degree of signal, and displayed. That is, "intensity" is a relative value of light and shade, strong scheduling, and the like in an image. During direct processing of the absorbance of the XP image, it is expressed as "density" or "change in density (Δ density)". And, for convenience in image representation, it is converted as described above and represented as "intensity". For example, when color display is performed with 256 gradations of 0 to 255, the display is "intensity". Such a distinction in terms applies in the case of XP or CT.

On the other hand, in the case of MRI, even if the air is specified as "-1000", the water is specified as "0", and the bone is specified as "1000", the values may significantly vary depending on the pixel value of MRI, the type of measurement device, the physical condition, physical form, and measurement time of a person at the time of measurement, and the acquisition method of the signal of MRI such as T1 emphasized image is various and not fixed depending on the facility and the type of measurement device. Therefore, in the case of MRI, "density" as in the case of XP or CT cannot be defined. Therefore, in MRI, relative values are processed from the stage of the initial drawing, and are expressed as "intensity" from the beginning. The processed signal is then also "intensity".

Through the above, basic data can be obtained. The basic data is extracted from a new object to be measured by using a waveform of the basic data, a certain fixed width of a wave Hz, and a certain fixed range of the wave Hz. For example, the extraction is performed by using only breath extraction, or by using the width, range, and waveform element of the degree of blood vessel extraction. The waveform and the width of Hz are determined relatively or comprehensively based on statistics using waveform elements of other functions, "artifacts" such as noise, waveforms of other "modes" in which other adjustability is considered, reproducibility of a plurality of times, and the like. Here, adjustment, experience (machine learning can also be applied) is required. This is because when the width and the range are wide, elements of other functions start to appear, and when the width is too narrow, elements of the function itself are eliminated, and therefore, the section needs to be adjusted. For example, when there are a plurality of times of data, it is easy to define a section, a width matching the Hz measurement, and the like.

[ concerning the adjustable coincidence rate ]

In this specification, the tendency of image change is described as an adjustable matching rate. For example, a lung field is detected, divided into a plurality of block regions, and the "average density (pixel value x)" of the block regions in each frame image is calculated. Then, a ratio (x') of "a change width (0% to 100%) from the minimum value to the maximum value of the average pixel value of the block region in each frame image" is calculated. On the other hand, only the block region having the ratio (x '/y ') within a predetermined fixed range is extracted using the value (x '/y ') of the ratio thereof to the ratio (y ') of the change (y) of the diaphragm of each frame image with respect to the change width (0% to 100%) of the minimum position to the maximum position of the diaphragm.

Here, in the case of y '═ x' or y ═ ax (a is a coefficient of the numerical value of the amplitude of the diaphragm or the numerical value of the "density"), both are completely identical. However, rather than having meaningful values only in the case of perfect agreement, values with some fixed width should be extracted. Therefore, in one embodiment of the present invention, a logarithm (log) is used to determine the fixed width as described below. That is, when the calculation is performed at a ratio (%) where y ═ x, it can be adjusted to "log y '/x' ═ 0" in full agreement. In addition, when the range of the extracted adjustable coincidence rate is a narrow (numerically narrow) range, for example, "log y '/x' — 0.05 to + 0.05" is determined in a range close to 0, and when the range of the adjustable coincidence rate is a wide (numerically wide) range, for example, "log y '/x' — 0.5 to + 0.5" is determined in a range close to 0. That is, the value of the logarithm of the adjustable rate is determined as a fixed range including 0. The smaller the range is, and the higher the value of coincidence in the range is, the higher the coincidence ratio can be said to be.

If the ratio is obtained for each pixel of the pixels and the number is counted, a normal distribution in which the peak is a case where the ratio is completely uniform can be obtained in the case of a healthy person. In contrast, in the case of a person with a patient, the distribution of the ratio collapses. As described above, the method of determining the width using the logarithm is merely an example, and the present invention is not limited thereto. That is, the present invention is directed to "image extraction" by providing (an average of "density" variation in a certain approximate range) about (movement of diaphragm) about (lung function inspection) about (area and volume of the lung channel), and can also apply a method other than the method using logarithm. By such a method, an adjustable image can be displayed.

In the case of a blood vessel, since there is a slight time delay (change in phase) while maintaining the shape in a series of changes in "density" (X (one waveform of the pulmonary gate portion)) occurring in concert with a series of contractions (Y) of the heart, Y is expressed as a' (X-t) (i.e., Y ≈ X). In the case of complete agreement, t is 0, and thus y is x or a' x. As in the case of the diaphragm, when the range of the extracted adjustable coincidence rate is a narrow (numerically narrow) range, for example, "log y '/x' — 0.05 to + 0.05" is determined in a range close to 0, and when the range of the adjustable coincidence rate is a wide (numerically wide) range, for example, "log y '/x' — 0.5 to + 0.5" is determined in a range close to 0. The smaller the range is, and the higher the value of coincidence in the range is, the higher the coincidence ratio can be said to be.

In the case of other blood vessels, the "portion corresponding to the heart" is excluded, and the "density" on the central side drawn according to the lung portal is used. In the case of peripheral blood vessels, the treatment can be performed in the same manner.

In addition, the present invention is also applicable to a circulator, for example, a change in "density" of the heart is directly related to a change in "density" of a blood flow flowing to the pulmonary portal section to the peripheral lung field, and a series of changes in "density" of the heart and changes in "density" of the pulmonary portal section are transmitted while being transformed and maintained. This is because it is considered that a plurality of phase differences are generated due to the relationship between the change in the "density" of the heart and the change in the "density" of the lung portal portion. Further, since the change in the "density" of the lung gate portion or the like is associated with the change in the "density" of the blood flow flowing to the lung field with the blood flow kept constant, the adjustability can be expressed by a case (the relationship of the matching rate of Y ≈ X) reflected on the rate with the blood flow kept constant. In addition, in the neck vascular system or the great vessel system such as the chest, abdomen, pelvis, and limbs, similarly, the change in "density" drawn with the heart vessels of the central center in the vicinity is directly related or related with a slight phase. Then, if the "density" fluctuates depending on the background and the state of the change in the "density" is transmitted during transmission, it can be considered as an adjustable coincidence rate.

Here, the "amount of inspiration totals about the amount of expiration" may be assumed to be 1 image and 1 image. Therefore, when a numerical value is calculated relatively from the difference in permeability with respect to the ambient air, if it is desired to display the numerical value as a relative value when the variation is set to 1 from the "density" of the lung field (Standard Differential signaling density/Intensity: Standard Differential signal density/Intensity), the following are calculated: (1) an image when 1 is set for each 1 image among different images for each 1 image (normally assumed); (2) a ratio in which 1 is an absolute value of an inhalation or exhalation with "density (change amount or change rate)" added to each of 1 image; in addition, (3) when the total amount of "density" in each breath (when 10% is selected several times) in the plurality of shots is 1, the change amount and the change rate can be plotted.

In addition, although there is a case of 3D such as MR, the difference between "intensity" (in the case of MR) and "density" (in the case of CT) of the whole inspiration may be converted into "peak flow volume deta" of inspiration (whether quiet or hard breathing) by adding up the values (in this case, when the values are set to 1), and the ratio of "intensity" or "density" may be calculated for the values, so that the actually measured respiration volume and the respiration rate of each lung field portion may be converted when at least "3D × time" in MRI or CT is calculated. Similarly, by inputting the cardiac output 1 time, the distribution of the "capillary phase" of the "flow" in the lung field can also present a distribution converted into the pulmonary blood flow peripheral amount and an estimated value of the volume.

That is, the (amount of change in inspiration per 1 picture) × (total number of inspired) x (amount of change in expiration per 1 picture) × (total number of expired breath) is approximately equal to (the volume of inspired breath: natural respiration or effort respiration at this time) ≈ about (the volume of expired breath: natural respiration or effort respiration at this time) ≈ about (the amount of change in inspiration or expiration of "volume of natural respiration or effort respiration at this time). When only the amount of change of 1 sheet of 10% or 20% is taken out, the estimated value can be calculated by calculating (the total number of sheets) × (the amount of change over time).

The extracted variation is visualized and rendered into an image. This is a respiratory function analysis and a blood vessel analysis described below. The rate of change of the thorax, diaphragm, etc. is then visualized. In this case, the artifact in the result may be eliminated again, and a waveform extracted from new data, a data waveform as the first base, a waveform of another mode or the like, a surrounding waveform, or a plurality of waveforms may be extracted to extract a function. The method of eliminating the artifacts will be described later.

Further, the feature amount may be grasped even if the variation components extracted from other than the above-described extraction are excluded. For example, when grasping the motion of the abdominal intestinal tract, the influence of respiration and the influence of blood vessels are excluded from the abdomen to realize the extraction of the motion of the abdominal intestinal tract.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. Fig. 1A is a diagram showing a schematic configuration of a diagnosis support system according to the present embodiment. The diagnosis assisting system exerts a specific function by causing a computer to execute a diagnosis assisting program. The basic module 1 includes a respiratory function analysis unit 3, a pulmonary blood flow analysis unit 5, another blood flow analysis unit 7, a fourier analysis unit 9, a waveform analysis unit 10, and a visualization/digitization unit 11. The base module 1 acquires image data from a database 15 via an input interface 13. The database 15 stores images based on DICOM (Digital Imaging and communication in Medicine), for example. The image signal output from the base module 1 is displayed on the display 19 via the output interface 17. Next, the functions of the basic modules according to the present embodiment will be described.

[ analysis of the period of respiratory elements ]

In the present embodiment, the period of the respiratory element is analyzed based on the following index. The "respiratory component" is a concept including all or a part of exhalation or inhalation as described above. That is, the frequency of at least one of the respiratory elements is analyzed using at least one of the "density"/"intensity" in a certain fixed region within the lung field, the motion of the diaphragm, and the motion of the thorax. In the "frequency of at least one of the respiratory elements", a frequency spectrum indicated by the respiratory element is one or more, and is a concept including a case having a fixed bandwidth. Since the lung fields are regarded as an aggregate of blocks, and a plurality of frequencies are extracted from each block, these frequencies are handled as a frequency group in the present embodiment. As described above, the basic data has both concepts of "the form of the wave" and "the interval of the wave (frequency: Hz)", and therefore can be handled as the form of the wave. Further, a range constituted by a certain fixed volume "density"/"intensity" measured at a site where the permeability of X-rays (other types of modalities such as CT and MRI) is high, data obtained by other measurement methods such as a spirogram, and externally input information may be used.

Further, the analysis results for each breath may be compared, and the trend may be analyzed from a plurality of data, thereby improving the accuracy of the data.

Further, the respiratory elements may be corrected by changing the phase of the frequency of at least one of the respiratory elements or smoothing the waveform of the respiratory element. In this case, the phase is combined with the wave using the motion (the motion of the thoracic cage, other diaphragm) about (density) about (the motion of the thoracic cage) about (the exact lung function) or the like. Further, the "density" of the lung field average is tracked, and for the final change, approximation such as binary multiplication of waves is performed as the wave form, and the waves are identified. Here, in the case of "density" of the chest or the like, the value with the largest change is the "density" of the lung, and therefore, the change in the "density" of the lung may be evaluated by evaluating the "density" of the entire screen. When plotting the wave, it is sometimes the case that the wave is actually moving, and a phase shift is sometimes generated in the measured value. In this case, the phase may be corrected at a position where the phase difference is maximum or minimum, or at the entire wave form.

[ waveform analysis ]

The frequency of the waveform can be calculated from the waveform of the respiratory component. Thereby, the above-described "waveform tunability image" is acquired. Specifically, a waveform of any portion within an analysis range is determined, a structural element of the frequency of the determined waveform is extracted, and an image corresponding to the structural element of the frequency of the waveform is output.

[ cardiovascular pulsation analysis and vascular pulsation analysis ]

In the present embodiment, the cardiovascular pulsation analysis and the blood vessel pulsation are analyzed based on the following indexes. That is, the heart, the lung portal position, and the main blood vessel are identified from the measurement results of other modalities such as an electrocardiogram and a pulsometer, and the change in "density"/"intensity" of each part is used to analyze the blood vessel pulsation. Further, the change in "density"/"intensity" of the target region may be analyzed by manually drawing the image. Then, a frequency (waveform) of at least one of the cardiovascular beat elements obtained from the heart beat or the blood vessel beat is determined. In addition, it is desirable to compare the analysis results of each beat and analyze the tendency from a plurality of data, thereby improving the accuracy of the data. Further, the accuracy can be improved by extracting the "density"/"intensity" of each region a plurality of times or extracting the "density"/"intensity" of each region for a fixed range. In addition, there are methods of inputting cardiovascular beat frequencies or frequency bands.

[ identification of the lung field ]

Images are extracted from a Database (DICOM), and the lung contour is automatically detected using the results of the periodic analysis of the respiratory elements. For the automatic detection of the lung contour, a conventionally known technique can be used. For example, the techniques disclosed in Japanese patent laid-open publication No. 63-240832 or Japanese patent laid-open publication No. 2-250180 can be used. Next, the lung fields are divided into a plurality of block regions, and the change of each block region is calculated. Here, the size of the block area may also be determined according to the shooting speed. When the imaging speed is slow, it is difficult to specify a corresponding portion in a frame image next to a certain frame image, and therefore the block area is increased. On the other hand, when the imaging speed is high, the number of frame images per unit time is large, and therefore tracking can be performed even if the block area is small. The size of the block area may be calculated based on which timing is selected in the period of the respiratory element. Here, it is sometimes necessary to correct the deviation of the lung field region. At this time, the motion of the thoracic cage, the motion of the diaphragm, and the positional relationship of the blood vessels of the entire lung field are recognized, and the relative position of the lung contour is grasped, and relative evaluation is performed based on the motion. Further, if the block area is too small, flickering of an image may occur. To prevent this, the block region needs to have a certain size.

In the above-described automatically detected lung field region, at least one bezier curve may be used to represent the lung fields as coordinates of points and control points. Then, the lung fields can also be represented by using a plurality of closed curves surrounded by at least one bezier curve, so-called "simple closed curves". Likewise, one or more simple closed curves can also be used to represent the analysis object.

The lung fields in the other frames may be detected using at least one Bezier curve (Bezier curve) on the lung fields detected in a specific frame. For example, a method of detecting the maximum and minimum 2 lung fields and calculating the lung fields of other frames using the 2 lung fields is cited. Here, the variable "change rate" is defined in other frames. The "change rate" is a value indicating the size of the lung field, that is, the state of breathing, and can be calculated from the position of the diaphragm, the "intensity" average value of the entire image, or the like. It is also possible to use the measurement data of an external device such as a spirograph to calculate or use a modeled rate of change. Since the variable "rate of change" can be arbitrarily determined in this manner, for example, it is also possible to calculate the lung fields assuming that the lung fields change at a fixed rate (10%, 20%, 30% …). Since the change rate defined in this way may include an error, the subsequent processing may be performed using a result obtained by performing automatic/manual elimination of the error, a result obtained by performing approximation using a least square method or the like, or the like. Assuming that linear deformation occurs until the maximum lung field and the minimum lung field are caused, the lung fields in each frame are calculated by a technique such as linear transformation using the change rate of each frame.

Further, the above-described processing can be applied to any range of consecutive frames. For example, although the lung field repeatedly changes to a maximum and a minimum in breathing, the shape of the maximum is not always constant in actual measurement. For example, in each range of extremely large to extremely small and extremely small to extremely large, it is desirable that the lung fields can be calculated with higher accuracy than defining and calculating the 2 lung fields of the maximum and minimum by applying the above-described processing. In addition, although the description has been made using the maximum and minimum as a specific example, the present invention is not limited to this, and the "arbitrary range" is used, and therefore, for example, the present invention can be performed at positions such as 0% and 30%, 30% and 100% in the middle of breathing.

Although the accuracy is lowered, the lung fields of each frame can be calculated from 1 lung field. For example, by analogy with the shape of the thorax, the change vector of the lung field is defined. Specifically, although the provision of the variation vector at each of the control points of the bezier curve is employed, the present invention is not limited thereto. Then, the lung fields in each frame are calculated using the detected 1 lung field and change vector, and the change rate in each frame. The accuracy can be further improved by correcting the calculation result in an automatic or manual manner. In addition, the above method is also effective in 3D. That is, even in the case of 3D, the process of detecting the lung fields in other frames can be executed using at least one bezier surface (beziersource) on the lung fields detected in a specific frame. This makes it possible to obtain an image of the lung fields between frames.

Fig. 6C shows a graph of the period of the respiratory element. Although the white vertical line is shown in the image of fig. 6C, this is a straight line (index) indicating the position at the current time in the period of the respiratory element, and the current position in the period of the respiratory element moves as the lung video shown in fig. 6B moves. By indicating the current position of the cycle of the respiratory element, the current position can be clearly grasped in the cycle of the movement of the lung. In the present invention, not only the period of the respiratory element is represented by a graph, but also all of the "density" of the blood flow, the thoracic cage, the diaphragm, and other parts linked to the movement of the lung can be represented by a graph.

In addition, in the case of the "stop breathing" of the subject, the frequency of the breathing component may not be specified. In this case, the fourier analysis described below is performed using the frequency of at least one of the cardiovascular pulsation elements extracted from the heartbeat or the blood vessel pulsation of the subject. In this case, the method of dividing the block region described below may be changed according to the motion pattern of the heart, the diaphragm, or the moving portion in conjunction with respiration.

[ detection of edge and evaluation thereof ]

The present invention is able to detect the edges of the lung and evaluate the edges. For example, after the lung fields are calculated by the aforementioned method, the position and shape of the edge can be detected again with high accuracy. A point is drawn at an arbitrary position within the calculated lung field, a line is extended from the point in all directions, and the change in pixel value in each line is evaluated. For example, as shown in fig. 14, when the pixel value is calculated along a line segment S cutting off the lung, although it is known that the pixel largely fluctuates at the edge, the absolute value of the fluctuation is different. For example, the accuracy of edge detection is improved by adjusting the threshold values for detecting the left edge and the right edge. Further, the characteristic of the pixel value variation for each region may be used. As shown in fig. 14, even if the difference between the edges of the S2 region and the S3 region is small, the edges of the S2 region and the S3 region can be specified from the variance of the fluctuation of the pixel value. Here, although the variance is focused on, the present invention is not limited thereto.

In addition, by the same consideration, the edge of the analysis range of organs other than the lung, blood vessels, tumors, and the like can be detected. For example, in the case where a contrast medium is present in a blood vessel, although the inside of the blood vessel can be clearly visualized, it is not easy to clearly calculate the outside and thickness of the blood vessel. In the present embodiment, since the edge can be accurately detected, the shape and the feature of the blood vessel within the analysis range can be calculated. This makes it possible to quantitatively determine the thickness and the outer periphery of a blood vessel, which have been difficult to determine in the past, and use the thickness and the outer periphery for diagnosis.

[ Generation of Block region ]

A method of dividing the lung field into a plurality of block regions will be described. Fig. 1B is a diagram illustrating a method of radially dividing lung fields from the pulmonary portal. In the lung, since the diaphragm side moves more greatly than the lung tip side, points that are more coarsely divided closer to the diaphragm side may be drawn. In fig. 1B, a line (broken line) for drawing the vertical direction may be added and divided into a plurality of rectangular (square) block regions. This enables the lung operation to be more accurately represented. The lung field can also be divided by a method such as "a method of drawing a point in the longitudinal direction of the lung and dividing the lung transversely", "a method of drawing a point in the transverse direction of the lung and dividing the lung longitudinally", "a method of drawing a tangent line at the tip of the lung and a tangent line at the diaphragm, and dividing the lung by a line segment drawn at a certain angle from a straight line (e.g., a vertical line) including the point with the intersection of the tangent lines as a center point", "a method of cutting the lung by a plurality of planes orthogonal to a straight line connecting the ends of the diaphragm from the tip (or the hilum of the lung)", and the like. In addition, these methods can also be applied to three-dimensional stereoscopic images. In the case of 3D, each organ is captured as a space surrounded by a plurality of curved surfaces or planes. The organ can be further finely divided. For example, when a 3D model of the right lung is considered, the processing may be performed in a manner divided into upper, middle, and lower lobes.

In the lung field region, the relative position of the lung contour is grasped by recognizing the positional relationship between the motion of the thoracic cage, the motion of the diaphragm, and the blood vessels of the entire lung field, and the relative evaluation is performed based on the motion. Therefore, in the present invention, after the lung contour is automatically detected, the region specified by the lung contour is divided into a plurality of block regions, and the values (pixel values) of the changes in the image included in each block region are averaged. For example, as shown in fig. 10, it is also possible to draw points on the opposite lung edges on the bezier curve, connect them, and then use a curve passing through these intermediate points. As a result, as shown in fig. 1C, even if the lung morphology changes with time, the time-dependent change of the region of interest can be tracked. On the other hand, fig. 1D is a diagram showing a temporal change in the case of dividing the target organ (lung in this case) into block regions, without considering the form of the target organ. The lung field region is a region to be relatively evaluated based on the motion of the thorax, the motion of the diaphragm, the positional relationship of the blood vessels of the entire lung field, and the grasp of the relative position of the lung contour, as described above, but if the lung field region is not determined by being divided into block regions as shown in fig. 1D, the region of interest deviates from the lung field region due to the temporal change of the lung, and becomes a meaningless image. In particular, since the lung fields are greatly contracted by the movement of the diaphragm, it is preferable to correct the lung field region by acquiring not only the diaphragm and the entire data but also the thoracic cage component and a plurality of other elements. Further, there is also a method of inputting a respiratory element frequency or a frequency band. For 3D, region segmentation may also be computed similarly.

As shown in fig. 11, in the lung field a, the internal control points may be selected in the detected lung fields using bezier curves, and the lung fields may be divided by curves or straight lines passing through the internal control points in the lung fields. That is, control points are provided not only in the frame of the lung field but also in the interior of the lung field region, and the lung field region (a) is divided using these control points. In this case, as shown in fig. 12, it is also possible to make: the intervals between control points on the extension of the detected lung fields and in the vicinity thereof are made relatively large (1), and the intervals between internal control points are made relatively small (2) based on the expansion ratio of each part in the detected lung fields. In the lung field a, the intervals of the control points may be made relatively large as they go forward in the head-tail direction with respect to the human body, or may be made relatively large as they go in a specific vector direction. The method of specifying the vector is arbitrary, and for example, the vector may be specified in a direction from the apex of the lung to the opposite side of the lung field, or may be specified in a direction from the portal of the lung to the opposite side of the lung field as shown in fig. 1B. Furthermore, the vector can also be determined in a direction corresponding to the structure of the lung. In this way, by setting the method of dividing the lung fields to "unequal division", an image in which the features of each region are taken into consideration can be displayed. For example, since the outer periphery of the lung fields moves largely and the deviation becomes large, the block becomes large, and the movement of the inner portion of the lung fields becomes small and the deviation becomes small, so that the block becomes small and fine. Further, for example, since the movement of the diaphragm side of the lung field is large and the deviation becomes large, the movement of the head side of the lung field is small and the deviation becomes small while the mass is made large, and thus the mass becomes small and fine. This improves the display accuracy. This method is not limited to the lung field, and can be applied to a moving part or the like linked with breathing. Such a method can also be applied to the case of 3-dimensional segmentation of the lung for each lobe. Further, the present invention can also be applied to a case where a part below the diaphragm, for example, the heart or another organ is surrounded by bezier curves and displayed. In this case, the vector can be determined in a direction corresponding to the structure of the heart or other organs, and the region can be divided unequally.

Then, the artifact is removed and the image data is interpolated. That is, if a bone or the like is included in the analysis range, it is represented as noise, and therefore, it is preferable to remove the noise using a denoising filter. In the X-ray image, in a general example, air is set to-1000 and bones are set to 1000, so that the pixel value of the portion with high transmittance is low and displayed in black, and the pixel value of the portion with low transmittance is high and displayed in white. For example, when a pixel value is expressed by 256 gradations, black is 0 and white is 255. In the lung field region, since X-rays are easily transmitted around a position where blood vessels and bones are not present, the pixel value of the X-ray image becomes low, and the X-ray image becomes black. On the other hand, at a position where a blood vessel or a bone exists, since the X-ray is hard to transmit, the pixel value of the X-ray image becomes high, and the X-ray image becomes white. The same can be said for other CT and MRI. Here, it is possible to eliminate artifacts by interpolating data using values of the same phase based on a waveform for each breath from the result of the period analysis of the respiratory element. Further, when "the coordinates are different", "the pixel value is extremely fluctuated", and "the frequency and the density are abnormally high" are detected, they may be deleted, and the continuous and smooth waveform may be recognized by using, for example, the least square method or the like for the remaining obtained images, so that it is possible to use for the calculation of the Hz of the diaphragm and the adjustment of the lung field. Further, in the case of superimposing images, there are: (1) a case where an acquired comparison image obtained by acquiring one of the images before and after is superimposed with its coordinates being maintained, and (2) a method of acquiring the one of the images before and after as a basis, relatively expanding the image, and superimposing the relative position information on the basis. With the above method, the form of the lung fields can be corrected, or the change in the image of the block region can be corrected. At this time, the artifact (artifact) corresponding to the result is removed again, and a waveform, a data waveform as the first base, a waveform of another mode or the like, a surrounding waveform, or a plurality of waveforms is extracted from new data to extract a function. In this case, the number of times may be one or more.

Here, "reconstruction" on the time axis is explained. For example, when the aspiration time is 2 seconds at 15f/s, 30+1 images can be obtained. In this case, if each 3 sheets are simply superimposed, every 10% of "reconstruction" can be implemented. At this time, for example, in a case where the 0.1 second is 10% and only 0.07 second and 0.12 second photographs are taken for the image, 0.1 second of "reconstruction" is required. In this case, the median (average of both) values of the images before and after 10% are provided to perform "reconstruction". Further, it is also possible to capture on a time axis and change the coefficient in proportion to the time. For example, when there is a difference in the time axis, there is no value of 0.1 second of shooting, there are shooting times of 0.07 seconds and 0.12 seconds, it may be recalculated to "(this 0.07 second value) × 2/5+ (0.12 second value) × 3/5" to perform "reconstruction". The change position relationship in the second is identified from the average of respiration and the amount of change in the coefficient of the diaphragm, and the ratio of the number is determined using the value as a coefficient. In addition, the thickness is preferably calculated to include 0 to 100% of the "maximum difference intensity projection", and to have a thickness of 10% to 20% of "reconstruction", or 10% to 40% of "reconstruction". In this way, even for a portion not photographed, "reconstruction" can be performed at a one-breath rate. In the present invention, the "reconstruction" can be similarly performed not only for respiration but also for blood flow, motion of the thorax, the diaphragm, and a series of other motions linked to these motions. Furthermore, "reconstruction" can also be performed for each block or for each pixel. In addition, the thickness is preferably calculated to include 0 to 100% of the "maximum difference intensity projection", and to have a thickness of 10% to 20% of "reconstruction", or 10% to 40% of "reconstruction".

Alternatively, the lung fields may be detected by the above-described method, and the detected lung fields may be normalized. That is, the detected lung fields are normalized spatially, or the detected lung fields are normalized temporally by reconstruction (reconstruction). The lung fields vary in size and shape depending on the human body, but can be displayed in a fixed area by standardizing them.

[ diaphragm and thorax ]

If the lung field is recognized as described above, the motion of the diaphragm and the thorax can be grasped. That is, the localization of the function evaluation from the image can be performed by calculating a curve obtained by the diaphragm on Xp (2D image) of the identified diaphragm or a curve of the thoracic cavity as a set of fine coordinates, and performing "curve fitting" using the average thereof, the local downward change rate or amount of the curve, and the diaphragm as curves to convert the deformation rate into numerical values. In addition, the function evaluation from the image may be performed by calculating a curve of the edge drawn on the chest other than the diaphragm surface as a set of fine coordinates and digitizing the average or the change rate of the curve. The 2 change rates and changes described above are evaluated as relative and mutual linkage, and the functions of the motion (movement) are evaluated by digitizing and imaging the change rates (such as the portions that are linked but not moving) differently.

Here, the "diaphragm and thoracic evaluation method" will be explained. First, the motion of the diaphragm is displayed with a left-right horizontal line orthogonal to the axis of the body (so-called median line) as an axis. Next, the line of the diaphragm was flattened as a baseline. I.e. to match the line of the diaphragm with a horizontal straight line. Then, the movement of the diaphragm was evaluated. The line of the diaphragm was stretched and flattened, and the orthogonal motion of the curve was evaluated. Next, the motion is evaluated on the outside of the thorax using a line connecting the apex of the lung to the thoracic angle of the diaphragm as a baseline (as an axis). The motion was evaluated by flattening the line of the thorax as a baseline, i.e., matching the line of the thorax to the line of the "lung apex-rib diaphragm angle". In addition, the line of the thorax was extended to the baseline and flattened, and the orthogonal motion of the curve was evaluated. Then, the curvature and the radius of curvature of the thoracic and diaphragm lines were evaluated. Then, the change is calculated as a "change amount", and the change amount is differentiated and evaluated as a "change rate".

Fig. 6B and 6C are diagrams showing an example of an image displayed on the display. In fig. 6B, the motion of the left lung is displayed as a video. In the image of fig. 6B, although a white horizontal line is shown, this is a straight line (index) representing the position of the diaphragm, and if the video is played back, it moves up and down following the movement of the diaphragm. In this way, by detecting the diaphragm and indicating the white horizontal line indicating the index indicating the position of the detected diaphragm, that is, the position of the diaphragm, it is possible to perform image diagnosis by a doctor. In addition, not only a part of the diaphragm but also all points can be identified by identifying the lung field-diaphragm line, so that diagnosis can be performed on one region of the diaphragm, such as the left and right sides and the outer and inner sides, or on the entire diaphragm. Similarly, not only the diaphragm but also the motion of the moving part in conjunction with breathing, for example, the motion of the thorax, can be determined from a line such as a tangent position or a line of the thorax recognized by the lung field. In this manner, in a case where the edge is assumed to move, the edge can be detected by acquiring the difference in the continuous images. For example, in most cases, a tumor is hard, while its surroundings are soft. Therefore, the tumor does not move so much, but the periphery thereof moves actively, and thus the edge of the tumor can be detected by obtaining the difference.

In 3D images such as MRI and CT, the surface of the diaphragm may be captured as a single coordinate or a three-dimensional curved surface, the coordinate or curve may be calculated as a set of fine coordinates (a set of a contour, a plane, and coordinates of the edge of the diaphragm), and the average or local downward rate of change or amount of change of the curve, and the diaphragm may be digitized by "curve fitting" using the rate of change as a curved surface to locate function evaluation from the image. In addition, the curved surface of the edge drawn on the chest other than the diaphragm surface may be similarly calculated as a set of fine coordinates, and the average or the rate of change of the curved surface may be digitized to evaluate the function from the image. The 2 change rates and changes described above are evaluated as relative and mutual linkage, and the functions of the motion (movement) are evaluated by digitizing and imaging the change rates (such as the portions that are linked but not moving) differently.

[ Fourier analysis ]

Based on the cycle of the respiratory element and the blood vessel pulsation cycle analyzed as described above, fourier analysis is performed on the value of "density"/"intensity" of each block region or the amount of change thereof. Fig. 2A is a graph showing the "intensity" change of a specific block and the result obtained by fourier analysis thereof. Fig. 2B is a diagram showing fourier transform results obtained by extracting frequency components close to the heart beat and changes in "intensity" of the frequency components close to the heart beat obtained by performing inverse fourier transform on the fourier transform results. For example, if the "intensity" variation of a particular block is fourier transformed (fourier analysis), the result as shown in fig. 2A can be obtained. Then, if a frequency component close to the heart beat is extracted from the frequency components shown in fig. 2A, the result shown on the right side with respect to the paper surface of fig. 2B can be obtained. By taking an inverse fourier transform of this, as shown on the left side with respect to fig. 2B, a "strength" variation adjustable with the variation of the heart beat can be obtained.

As shown in fig. 9, a specific spectrum can be weighted by multiplying it by a coefficient. For example, to achieve waveform tunability, this approach can be used. That is, as a method of selecting a frequency when performing the inverse fourier transform, a plurality of frequencies are selected and multiplied by the ratio, and then the inverse fourier transform is performed. For example, if it is desired to highlight the spectrum with the highest frequency in the extracted frequency band, the spectrum intensity can be increased by 2 times. In this case, the continuity of the frequency is not concerned. The frequency spectrum in which the discontinuity exists can be selected.

The position of the "density" of the heart can be inferred from the morphology (the region of the left lung depressed portion in the morphology extracted from the lung field) of the left lung (in the case of the right heart, for example, in the case of the visceral inversion) and the positions of the vertebral body and the diaphragm. In this case, the ROI of the heart is acquired and the "density" extraction is performed. In this extraction, the rough region is used to perform analogy using the relative spectral values of respiration and blood flow. In addition, there are cases where the frequency due to respiration or other "artifacts" is removed by performing "filtering" in advance using an Hz band (heartbeat 40 to 150Hz, approximately equal to 0.67Hz to 2.5Hz) or the like generated by cardiovascular pulsation. Further, since the position of the heart also changes according to the breathing condition, the position of the heart may be relatively changed according to the morphological value of the thorax as the position of the thorax changes, and the extraction of the cardiovascular beat and the extraction of the pulmonary portal, the great vessel, and the like may be more accurately performed. In addition, as with the movement of the diaphragm, there is a method of calculating the frequency based on the contour of the heart that moves regularly.

Here, when performing inverse fourier transform on a frequency spectrum composed of frequency components, both frequency elements (respiratory frequency, cardiovascular pulsation frequency) and a frequency band of the frequency spectrum (BPF: band pass filter may be used) determined according to the "density" of respiration or blood flow may be added, or inverse fourier transform may be performed based on either one of them. Further, at least one frequency at the time of performing the inverse fourier transform may be selected based on a spectral composition ratio of a periodic change specific to the organ from among the frequency spectra obtained after the fourier transform. In addition, the waveform of a region to be a specific organ or an analysis target can be specified by the component ratios of a plurality of frequencies obtained after fourier transform (generation of a waveform tunability image).

In addition, in order to calculate the fourier transform in a short time, an AR method (Autoregressive Moving average model) may be used. In the AR method, in the autoregressive moving average model, there is a method using Yule-walker equation (Yule-walker equation), kalman filter, and Yule-walker estimates (Yule-walker estimates), PARCOR method, least square method derived therefrom may be used to supplement the calculation. This enables a near-real-time image to be acquired more quickly, and assistance in calculation and correction of an artifact (artifact) to be performed. By such fourier analysis, the properties of the image in each block region can be extracted and displayed.

In addition, a method using a "digital filter" can also be employed in the fourier analysis. That is, the original waveform is fourier-transformed to obtain parameters of each spectrum, and a "digital filter" that performs arithmetic processing on the original waveform is used. In this case, a digital filter is used without performing an inverse fourier transform.

Here, the fourier transform may be performed on the change in the image of each block region in each frame image, and the spectrum in the fixed band including the spectrum corresponding to the cycle of the respiratory element may be extracted from the spectrum obtained after the fourier transform. Fig. 2C is a diagram showing an example of extracting a certain fixed band from a spectrum obtained after fourier transform. The frequency f of the frequency spectrum of the synthesized wave is at each frequency f as the synthesis source₁(respiratory component), f₂(blood flow component) with the establishment of "1/f-1/f₁+1/f₂"such a relationship can be used in the following method when extracting the spectrum.

(1) The portion of the blood flow where the spectral ratio is high is extracted.

(2) The spectrum is extracted by dividing the peak of the spectrum corresponding to the respiration/blood flow and the middle of the peaks of the plurality of synthesized waves in the vicinity thereof.

(3) The peak of the spectrum corresponding to the respiration/blood flow and the valleys of the spectrum of the plurality of synthesized waves in the vicinity thereof are divided to extract the spectrum.

As described above, in the present invention, a spectrum in a fixed frequency band including a spectrum corresponding to a period of a respiratory element is extracted without using a fixed BPF. In the present invention, a spectrum in a fixed frequency band including a frequency (for example, a spectrum model) corresponding to a frequency (for example, "density"/"intensity" of each part, a heartbeat element obtained from a heartbeat or a blood vessel pulsation) obtained from a frame image other than a respiratory element or a frequency inputted from the outside by an operator can be extracted from a spectrum obtained after fourier transform.

Here, if the component of the frequency spectrum of the synthesized wave is only 2 components (respiration, blood flow), it is 50% + 50%, and in the case of 3 components, it is assigned 1/3 each. Therefore, the spectrum of the synthesized wave can be calculated to some extent from the components of the spectrum and the heights thereof, regardless of the spectrum of the respiratory component and the spectrum of the blood flow component. The spectrum can be extracted when the ratio (%) is high. That is, the ratio of the blood flow component/respiratory component to the composite wave is calculated, and the spectral value of the blood flow component/respiratory component higher is calculated and extracted. In the identification of the diaphragm, etc., a spectrum corresponding to a region where Hz (frequency) is relatively constant, that is, a region where Hz changes little, or a superposition thereof may be extracted from data obtained by acquiring the frequencies of the respiration and the cardiac blood vessels. When the frequency band of the spectrum is specified, the frequency band of the spectrum may be specified in a section where the change in Hz occurs and a region around the section when the identification of the diaphragm is performed. There are cases where a structural element of a waveform is added.

In addition, as for the spectrum at the time of the inverse fourier transform, "a case (meaning of simulation) in which a higher portion(s) is/are extracted from the simply modeled frequencies and frequency bands, and" a case (meaning of scene) in which a higher portion or a lower portion is extracted from the spectrum values based on the actual frequencies or frequency bands can be selected. When the frequency of the heart is a and the frequency of the lung is B, B can be obtained by subtracting a from the entire frequency band. Further, not only one position on the frequency axis but also a plurality of positions can be extracted for the spectrum obtained by fourier transform.

As described above, the frequency can be extracted not only when the frequency completely coincides with the period of the respiratory element and the blood vessel pulsation period, but also the frequency to be considered, and this can contribute to the image diagnosis. In addition, it is known that "respiration" or "heartbeat" is contained in a specific frequency band. Therefore, for example, a filter of "0 to 0.5Hz (respiration rate 0 to 30/min)" may be used for respiration, and a filter of "0.6 to 2.5 (heartbeat/pulse rate 36 to 150/min)" may be used for circulator, and the respiratory rate and the circulator rate may be determined in advance by the filter. This enables display of a frequency-adjustable image. This is because, when the "density" of the heart is acquired, the "density" of the respiration (lung) is picked up, or when the "density" of the lung is acquired, the "density" of the heart is picked up.

[ visualization and digitization ]

The results obtained by the analysis as described above were visualized and digitized. In the visualization and the digitization, a "modeled lung" is defined in the present specification. When the lung is displayed as a video image, the relative determination is not easy because the positional relationship moves. Therefore, the variation in the positional relationship is spatially normalized and averaged. For example, the shape of the lung is matched with a pattern such as a fan shape, and displayed in a trimmed state. The concept of reconstruction is then used to normalize over time. For example, the "state of 20% of lungs in a plurality of breaths" may be extracted and determined as "condition of 20% of lungs in one breath". In this way, the spatially and temporally uniform lung is referred to as a "modeled lung". This makes it easy to perform relative judgment when comparing different patients with each other or comparing the current and past of one patient.

For example, as a standard uptake (standard uptake), a value may be displayed in a relative/logarithmic manner with an average value of 1 being set from the measured "density"/"intensity" of the entire lung field region. Further, since only the direction of blood flow is used, a change toward a specific direction is sometimes intercepted. This makes it possible to extract only meaningful method data. The lung field recognition result is used to perform pseudo-colorization following the change of the analysis range. That is, the analysis result of each individual (subject) is matched to the relative area based on the specific shape (minimum value, maximum value, average value, central value) matched to the stage.

Further, the analysis result is deformed into a specific shape and stage which can be compared with each other. In addition, when generating a modeled lung, the relative positional relationship in the lung field is calculated using the results of the period analysis of the respiratory elements. The modeled lung is generated using a line obtained by overall averaging the thoracic line, "density", diaphragm, and the like of a plurality of patients. In generating a modeled lung, the distance may be measured radially from the ostium to the end of the lung for the case of pulmonary blood flow. In addition, in the case of breathing, correction is required according to the movement of the thorax and diaphragm. Further, the comprehensive calculation may be performed in consideration of the distance from the lung apex.

After the inverse fourier transform, only a block having a relatively large amplitude value may be extracted and displayed. That is, when fourier analysis is performed for each block, there are a block in which the amplitude of the wave is large and a block in which the amplitude of the wave is small after the inverse fourier transform. Thus, it is also effective to extract only a block having a relatively large amplitude and visualize the extracted block. In addition, the real part and the imaginary part of each numerical value may be separately used after fourier transform, respectively. For example, the image can be reconstructed from only the real part, or from only the imaginary part, or from the absolute values of the real and imaginary parts.

Fourier analysis can also be performed on the modeled lung. That is, the modeled lung may be used also when images of a plurality of breaths are combined, fourier analysis is performed, and relative position is grasped. By matching the acquired plurality of frames with the modeled lung using the modeled lung, or in the case of a blood vessel, matching the acquired plurality of frames with the modeled lung calculated from a heart beat (for example, a heart beat obtained from the pulmonary portal part), it is possible to fix the relative position in the case of performing fourier analysis. Even when the state of the basic respiration is acquired, the calculation result can be stably obtained by using the modeled lung. Furthermore, by modeling the lung, spatial differences can be immobilized, and the movement of the lung can be easily observed.

In the imaging, the identification method of the relative evaluation is as follows. I.e. the images are relatively marked with a black-and-white, color mapping. Values around several% of "density"/"intensity" obtained by the difference may be removed, and the upper and lower remaining portions may be relatively displayed. Alternatively, since values before and after a few% of the obtained difference may be highlighted, the value may be removed as an "artifact" and the remaining portion may be relatively displayed. In addition to the method of 0 to 255 gradations, the display can be performed as a value of 0 to 100%.

In addition, the pixels can be displayed to some extent with blurring, and the entire display can be displayed in a blurred state. In particular, in the case of pulmonary vessels, although a low signal value is mixed between high signal values, if only a high signal value can be roughly grasped, there is no problem even if the whole is unclear. For example, a signal equal to or higher than a threshold value may be extracted for the case of blood flow, and a signal equal to or higher than a threshold value may not be extracted for the case of respiration. Specifically, when the numbers in the following table are set to 1 pixel and an intermediate numerical value is acquired, if the proportion occupied by the intermediate numerical value is acquired and averaged within 1 pixel, it can be smoothly expressed between adjacent pixels. This method can also be used in the calculation of the average intensity of each block.

[ Table 1]

This method can be applied not only to the lung field but also to the case where the density (density) in an arbitrary analysis range is detected and a position where the density changes relatively largely is removed. In addition, points that greatly exceed a preset threshold are deleted. In addition, the shape of the rib is recognized, for example, by recognizing and removing a high/low signal line which suddenly appears. In addition, as with the phase, a sudden signal different from a normal wave change, such as a sudden signal whose artifact is observed before or after the stage of reconstruction is 15% to 20%, may be removed. Further, in the initial acquisition of the basic data, there are cases where the phase is different in the calculation of (diaphragm) ≈ (thoracic operation) ≈ (spirometry) ≈ or (density measurement) in the region (field), or the like, and there are cases where the phase matches the pattern (profile of XP) actually recognized.

If a modeled lung can be generated, the adjustability, the coincidence rate, and the non-coincidence rate can be presented numerically (display of a frequency adjustability image or a wavelength adjustability image) as described above. Thereby, the case of the deviation from the normal state can be displayed. According to the present embodiment, by performing fourier analysis, it is possible to find a new possibility of a disease, compare with oneself in a normal state, compare a hand with a foot, and compare a hand and a foot on the opposite side. In addition, it is possible to grasp whether or not there is a lack of strength in the exercise mode, swallowing, and the like of the foot by digitalization of adjustability. In addition, it is possible to determine whether or not a person in a diseased state changes after a certain period of time has elapsed, and when the change occurs, compare the time before and after the change. Further, by setting the lung field to a form (circular to quasi-circular) in which the distance from the tip is constant and the lung field is easily observed radially, evaluation of the inner layer to the middle layer, the outer layer, and the like can be performed relatively easily, and the "peripheral superiority of blood vessels" or "middle superiority" can be expressed.

In addition, in the visualization, the image after the fourier transform and the image before the fourier transform can be displayed in a switched manner, or both can be displayed in a single screen.

As shown in fig. 2D, when the modeled lung is set to 100, it is possible to grasp whether there is a difference in percentage in the human body, and display the change rate. In addition, not only the difference of the whole lung but also the difference of a part of the lung can be grasped. In particular, as described above, it is possible to fix the shape of the lung field other than the diaphragm while only the movement of the diaphragm is determined, and to display the adjustable coincidence rate or the change rate while displaying the movement of the diaphragm. In addition, the entire lung fields can be fixed, and the adjustable matching rate or the change rate can be indicated. In addition, the standard blood flow can be determined by performing "change classification". That is, the cycle of the respiratory element can be specified, the relative positional relationship of the blood vessels can be calculated, and the blood flow dynamics of the subject can be specified as the standard blood flow.

In addition, a pattern matching method may also be used to detect the lung. Fig. 2E to 2H are diagrams showing examples of pattern images of the lung field region. As shown in fig. 2E to 2H, the lung shapes may be subjected to pattern classification in advance, and only the patterns close to these may be extracted. By this method it can be determined whether the image of the object represents a single lung or two lungs. In addition, it may also be determined whether the right or left lung. The number of patterns is not limited, and it is assumed that there are 4 to 5 patterns in advance. In addition, there is also a method of recognizing the right lung, the left lung, and both the lungs by using only the shape (shape) of the lung field. Further, a method of recognizing a thick band-shaped "low-permeability portion" due to the vertebral body and mediastinum, and recognizing the left and right lungs or both lungs based on a positional relationship with the band-shaped low-permeability portion and a positional relationship with the "high-permeability portion" of the lung field can also be employed. Furthermore, as shown in fig. 2H, the method may also be applied to the area below the diaphragm. This also allows the lower part of the diaphragm and the heart to be identified.

Further, since air has the highest permeability and is a portion having a higher permeability than the lung field, it is preferable to calculate the permeability by considering air as well. That is, the position of the air on the screen can be determined as follows.

In the case of (the area of air on the upper right of the screen) > (the area of air on the upper left of the screen), the left lung is identified. This is because the area of air outside the human body becomes wider in shooting for around the shoulder.

In the case of (the region of air on the upper left of the screen) > (the region of air on the upper right of the screen), the right lung is identified. This is also because, as described above, the region of air outside the human body widens in shooting for the surroundings of the shoulder.

Next, when (the region of the air on the upper right of the screen) is approximately equal to (the region of the air on the upper left of the screen), two lungs are recognized. This is because the air region is the same on the left and right.

Further, air in the intestinal tract may enter under the diaphragm, and at this time, it may not be recognized. Therefore, the most approximate lung field and its surrounding low-permeability portions such as the mediastinum side, the heart side, and the diaphragm side are recognized from the center of the lung field, and the line is recognized as the deep portion of the lung field. The method can also be used, for example

"https: (vii) com/help/images/examples/blocks-processing-big e-images _ ja _ jp.html ".

This allows a certain patient to be compared or quantified with other patients. In addition, the normal lung or normal blood vessels can be compared or quantified with typical abnormal lung function or abnormal blood flow. As a relative evaluation of lung function and pulmonary blood flow of a certain patient at different times, a modeled lung and a standard blood flow can be used. The lung and the standard blood flow thus modeled can be used as indexes for evaluating a certain patient in a morphologically matching manner by collecting typical patients of various types and typical examples of healthy persons and using the lung and the standard blood flow as modeled.

[ drawing of Lung fields ]

In general, since the lung field includes ribs having low permeability, it is difficult to mechanically recognize the outline of the lung using only "density" as an index. Therefore, in the present specification, a method is adopted in which the contour of the lung lobe is hypothetical-drawn using a combination of bezier curves and straight lines, and the lung contour is adjusted so that the consistency becomes high.

For example, if the contour of the left lung is represented by 4 bezier curves and 1 straight line, the lung contour can be drawn by obtaining 5 points and 4 control points on the lung contour. By drawing a plurality of lung contours with shifting the positions of points and evaluating the consistency using the conditions such as "the total value of the 'densities' in the contours is maximum" and "the difference between the total of the 'densities' of a plurality of pixels on the inner and outer sides of the contours is maximum", the lung contours can be detected with high accuracy. In practice, the positions of several points can be identified from the contour of the upper part of the lung where the edge is relatively easily detected and the position of the diaphragm detected by the method described later, and the number of attempts of the simulation can be suppressed. The position of the control point of the bezier curve can also be adjusted by extracting a contour using classical binarization, extracting a point near the outer edge, and using a least square method or the like.

Fig. 3A and 3B are diagrams showing an example in which lung fields are plotted using both bezier curves and straight lines. Fig. 3A shows a case where the lung area is the largest (maximum outline), and fig. 3B shows a case where the lung area is the smallest (minimum outline). In each figure, "cp 1 to cp 5" represent control points, and "p 1 to p 5" represent points on a bezier curve or a straight line. In this way, if the maximum contour and the minimum contour can be grasped, the maximum contour and the minimum contour can be obtained by calculating the contour in the middle. For example, the state of 10%, 20% … of exhalation can be displayed. As described above, according to the present embodiment, at least the lung field, the blood vessel, or the heart can be plotted using at least one bezier curve (bezier curve). The above method is not limited to the lung, and can be applied to other organs as "detection of an organ". Further, for example, on an analysis range (tumor, lower visual mound of brain, basal nucleus, boundary of inner capsule, etc.) predetermined in a specific frame, the following processing can be performed: at least one Bezier surface is used to detect the range corresponding to the analysis range in other frames.

Further, the present invention can be applied not only to a planar image but also to a stereoscopic image (3D image). By defining the equation of the curved surface and setting the control point thereof, the range surrounded by the plurality of curved surfaces can be set as the analysis target.

[ detection of movement of diaphragm or moving part in conjunction with respiration ]

In the images obtained by the continuous imaging, the movement of the diaphragm or the moving part linked with the breathing can be detected. If images are selected at arbitrary intervals among images obtained by continuous shooting and a difference between the images is calculated, the difference becomes large particularly for a region having a large contrast. By appropriately visualizing the difference, it is possible to detect the region where motion exists. In the visualization, continuity of a region having a large absolute value of the difference may be emphasized by using noise removal by a threshold value, curve fitting by a least square method, or the like.

In the lung field, the contrast of the line tangent to the diaphragm or the heart becomes significant, as shown in fig. 4A, and if the difference is taken in 2 lung images and a fixed threshold is set to visualize the difference, as shown in fig. 4B, the line tangent to the diaphragm or the heart can be visualized.

[ estimation of motion of diaphragm ]

In this method, although the diaphragm position can be detected when the diaphragm moves between the object images, it is difficult to detect a position where the movement of the diaphragm becomes slow. That is, it is difficult to detect immediately after the start of imaging or immediately before the end of imaging during a period when breathing is stopped at the timing of switching between expiration and inspiration. In the present method, the motion of the diaphragm is estimated using an arbitrary complementation method.

Using the method described above, after visualizing the diaphragm line, as shown in fig. 4B, the 1024 px-long image is divided into 128 rectangles per 8px long, the signal values included in the respective rectangular areas are summed up, and a histogram is plotted as shown in fig. 4C. Among the plurality of peaks, a peak closest to the lower coordinate indicated by a rectangle of a dotted line is expected to represent the position of the diaphragm. In a typical upright XP image, the diaphragm is shown as a curve, but the coordinates are approximated as the position of the diaphragm.

If the method is used to detect the diaphragm position for a full image, the "peak position" is detected as shown in FIG. 5. The motion of the diaphragm is estimated by correcting the detected value. First, when the difference is larger than the fixed value, it is regarded as a deviation value and excluded (a thin solid line in fig. 5). The data obtained after the elimination of the deviation values were divided into arbitrary clusters, 4-time curve regression was performed on each cluster, and the results were connected (thick solid line in fig. 5). In this analysis, although regression analysis is performed, the present invention is not limited to this, and any complementary method such as spline interpolation may be used.

[ refinement of moving part detection ]

There are cases where the contrast of the moving part varies along the line. In this case, the shape of the active portion can be detected more accurately by changing the threshold for noise removal and performing the detection process a plurality of times. For example, in the left lung, the contrast of the line of the diaphragm tends to become weaker as it enters the human body. In fig. 4B, only the right half of the diaphragm can be detected. At this time, by changing the setting of the threshold for noise removal, the remaining portion of the left half of the diaphragm can also be detected. By repeating this process a plurality of times, the shape of the entire diaphragm can be detected. With this method, the rate of change or amount of change in line or plane can be quantified for the shape as well as the position of the diaphragm, and this can contribute to a new diagnosis.

The position or shape of the diaphragm thus detected can be used for diagnosis. That is, in the present invention, coordinates of the diaphragm are graphed, and coordinates of the thorax or the diaphragm are calculated using the curve (surface) or the straight line calculated as described above, or "density" of the heartbeat, the blood vessel pulsation, the lung field, or the like can be graphed as a position or coordinates corresponding to the period. Such a method can also be applied to an active site in conjunction with breathing.

With this method, not only the Hz during inhalation and exhalation, but also when the frequency (Hz) of the diaphragm or the active site in conjunction with respiration changes, the measurement can be performed in a frequency band corresponding to the change. Then, when extracting the frequency spectrum of the BPF (bandpass filter), the BBF can be installed in a fixed range according to the state of each breath, and in the 'reconstruction stage' of each breath, the axis of the BPF position is changed to generate the optimal state, and the variable BPF matched with the optimal state is generated. This slows down the breathing, and even if there is a fluctuation in the rhythm of breathing, as in the case of a stop (Hz — 0), an image corresponding to the fluctuation can be provided.

Further, the frequency of the whole of expiration or inspiration may be calculated based on the ratio of the breathing element to the whole of expiration or inspiration. In the detection of the diaphragm, the detection may be performed a plurality of times, and the selection signal or the waveform may be stabilized. Based on the above, the frequency of at least one of the respiratory elements is calculated from the detected position or shape of the diaphragm or the position or shape of the active site linked with respiration. If the position or shape of the diaphragm or the active site can be grasped, the frequency of the respiratory element can be grasped. According to this method, even if a part of the waveform is divided, the waveform after that can be tracked. Therefore, even if the frequency of the respiratory component changes in the middle, the respiratory component can follow the original respiratory component. Although the pulsation of the heart and the like may suddenly change, the present invention can be applied to the cardiovascular system. Next, the operation of each module according to the present embodiment will be described.

[ analysis of respiratory function ]

First, the respiratory function analysis will be explained. Fig. 6A is a flowchart showing an outline of the respiratory function analysis according to the present embodiment. The basic module 1 extracts the DICOM image from the database 15 (step S1). Here, at least a plurality of frames of images contained in one respiratory cycle are acquired. Next, in each of the acquired frame images, the period of the respiratory element is specified using at least the density (density/intensity) in a certain fixed region within the lung field (step S2). In addition, the determined breathing cycle or a waveform determined from the breathing cycle can be used in the following steps.

In addition, the period of the respiratory element can be determined by using the motion of the diaphragm and the motion of the thorax. In addition, a range including a certain fixed volume and "density"/"intensity" measured at a site having high X-ray transmittance, or data obtained by other measurement methods such as a spirogram may be used. Note that the frequency of each organ (here, lung) may be determined in advance, and the "density"/"intensity" corresponding to the determined frequency may be extracted.

Next, in fig. 6A, the lung fields are automatically detected (step S3). Since the lung contour changes continuously, if the maximum shape and the minimum shape can be detected, the shape between them can be interpolated by calculation. The lung contour in each frame image is determined by interpolating each frame image based on the cycle of the respiratory element determined in step S2. Further, pattern matching as shown in fig. 2E to 2H may be performed to detect the lung fields. Further, noise removal by deletion may be performed on the detected lung fields. Next, the detected lung fields are divided into a plurality of block regions (step S4). Then, the change of each block region in each frame image is calculated (step S5). Here, the values of the variation within each block area are averaged and expressed as 1 data.

Further, noise removal by deletion may be performed for a value that changes in each block region. Next, fourier analysis or analysis of an adjustable matching rate is performed on the value of "density"/"intensity" of each block region and the variation thereof based on the cycle of the respiratory element (step S6).

Next, noise removal is performed on the result obtained by fourier analysis or analysis of the adjustable matching rate (step S7). Here, the deletion or the removal of the artifact (artifact) can be performed as described above. The operations of step S5 to step S7 are performed 1 or more times to determine whether or not the operations are completed (step S8). Here, since the characteristic amount displayed on the display is mixed with the synthetic wave or other waves, there is a case where a frequency modulation image of a high-purity element, for example, a respiratory element, a blood flow element, or other elements cannot be displayed by extracting the frequency spectrum once. In this case, the feature value displayed on the display may be used as a pixel value, and analysis may be performed again for all or a part of the display a plurality of times. By this operation, it is possible to further acquire an image with high purity relating to the adjustability or consistency of the factor, for example, the respiratory factor or the blood flow factor. In this operation, the operator may manually recognize the image on the display while visually recognizing the image, or may automatically extract the spectrum from the output result and recalculate the distribution ratio. Even after the calculation, the noise removal processing, the space filling (interpolation) by the least square method, and the correction using the "density" of the surroundings may be performed according to the case.

In step S8, if not completed, the process proceeds to step S5, and if completed, the result obtained by fourier analysis or analysis of the adjustable coincidence ratio is displayed on the display as a false-color image (step S9). In addition, black and white images may also be displayed. The accuracy of the data can also be improved by repeating a number of cycles in this manner. This enables display of a desired video. In addition, a desired video may be obtained by correcting an image displayed on the display.

In the present embodiment, the desired frequency or frequency band is calculated by calculation, but if viewed as an actual image, the display is not necessarily limited to a good image. Therefore, the following method may be used.

(1) A method of presenting a plurality of frequency bands and selecting the frequency bands by a person;

(2) a method of prompting a plurality of frequency bands and extracting a good image by an AI technique using pattern recognition;

(3) the selection is made according to the tendency and morphology of HISTGRAM. That is, the value of the center portion of the "histogram (Histgram)" in the resulting signal tends to be high, and the value of the "histogram" fluctuates according to the motion, and therefore, it can be selected according to the tendency and form of the Histgram.

[ analysis of pulmonary blood flow ]

Next, the analysis of the pulmonary blood flow will be described. Fig. 7 is a flowchart showing an outline of the pulmonary blood flow analysis according to the present embodiment. The base module 1 extracts the DICOM image from the database 15 (step T1). Here, at least a plurality of frame images included in one heart cycle are acquired. Next, a blood vessel pulsation period is determined based on the acquired frame images (step T2). In addition, the specified blood vessel pulsation period or a waveform specified from the blood vessel pulsation period can be used in the following steps. As described above, the blood vessel pulsation cycle is analyzed using the measurement results of other modalities such as an electrocardiogram and a pulsometer, and the change in "density"/"intensity" of an arbitrary portion such as the heart, the lung portal, and the main blood vessel. Note that the frequency of each organ (here, the pulmonary blood flow) may be determined in advance, and the "density"/"intensity" corresponding to the determined frequency may be extracted.

Next, in fig. 7, the period of the respiratory element is determined by the above-described method (step T3), and the lung field is automatically detected using the period of the respiratory element (step T4). In the automatic detection of the lung contour, although there may be a variation in each frame image, the lung contour in each frame image is specified by interpolating each frame image based on the period of the respiratory element specified in step T3. Further, pattern matching as shown in fig. 2E to 2H may be performed to detect the lung fields. Further, noise removal by deletion may be performed on the detected lung fields. Next, the detected lung fields are divided into a plurality of block regions (step T5). Then, the change of each block region in each frame image is calculated (step T6). Here, the values of the change in each block area are averaged and represented as 1 data. Further, noise removal by deletion may be performed for a value that changes in each block region. Next, fourier analysis or analysis of the adjustable matching rate is performed on the "density"/"intensity" value of each region and the variation thereof based on the blood vessel pulsation cycle (step T7).

Next, noise removal is performed on the result obtained by fourier analysis or analysis of the adjustable coincidence rate (step T8). Here, the deletion or the removal of the artifact (artifact) can be performed as described above. The operations of step T6 to step T8 are performed 1 or more times to determine whether or not the operations are completed (step T9). Here, since the characteristic amount displayed on the display is mixed with the synthetic wave or other waves, there is a case where a frequency modulation image of a high-purity element, for example, a respiratory element, a blood flow element, or other elements cannot be displayed by extracting the frequency spectrum once. In this case, the feature value displayed on the display may be used as a pixel value, and analysis may be performed again for all or a part of the display a plurality of times. By this operation, it is possible to further acquire an image with high purity relating to the adjustability or consistency of the factor, for example, the respiratory factor or the blood flow factor. In this operation, the operator may manually recognize the image on the display while visually recognizing the image, or may automatically extract the spectrum from the output result and recalculate the distribution ratio. Even after the calculation, the noise removal processing, the space filling (interpolation) by the least square method, and the correction using the "density" of the surroundings may be performed according to the case.

In step T9, in the event of incompletion, the process proceeds to step T6, and in the event of completion, the result obtained by fourier analysis or analysis of the adjustable coincidence rate is displayed on the display as a false-color image (step T10). In addition, black and white images may also be displayed. This can improve the accuracy of the data. In addition, a desired video may be obtained by correcting an image displayed on the display.

In the present embodiment, the desired frequency or frequency band is calculated by calculation, but if viewed as an actual image, the display is not necessarily limited to a good image. Therefore, the following method may be used.

(1) A method of presenting a plurality of frequency bands and selecting the frequency bands by a person;

(2) a method of prompting a plurality of frequency bands and extracting a good image by an AI technique using pattern recognition;

(3) the selection is made according to the tendency and morphology of HISTGRAM. That is, the value of the center portion of the "histogram" in the resulting signal tends to be high, and the value of the "histogram" fluctuates according to the motion, and therefore, the selection may be made according to the tendency and form of the histrgram.

[ other blood flow analysis ]

Next, another blood flow analysis will be described. As shown in fig. 15, one embodiment of the present invention can also be applied to blood flow analysis of the heart, aorta, pulmonary vessels, upper carpal arteries, neck vessels, and the like. Also, blood flow analysis can be performed on abdominal blood vessels, peripheral blood vessels, and the like, which are not shown. Fig. 8 is a flowchart showing an outline of another blood flow analysis according to the present embodiment. The base module 1 extracts the DICOM image from the database 15 (step R1). Here, at least a plurality of frame images included in one heart cycle are acquired. Next, a blood vessel pulsation period is determined based on the acquired frame images (step R2). In addition, the specified blood vessel pulsation period or a waveform specified from the blood vessel pulsation period can be used in the following steps. As described above, the blood vessel pulsation cycle is analyzed using the measurement results of other modalities such as an electrocardiogram and a pulsometer, and the change in "density"/"intensity" of an arbitrary portion such as the heart, the lung portal, and the main blood vessel. Further, the frequency of each organ (for example, main blood vessel) may be determined in advance, and the "density"/"intensity" corresponding to the determined frequency may be extracted.

Next, an analysis range is set (step R3), and the set analysis range is divided into a plurality of block regions (step R4). Then, the values of the changes within each block area are averaged and represented as 1 data. Further, noise removal by deletion may be performed for a value that changes in each block region. Next, fourier analysis or analysis of the adjustable matching rate is performed on the value of "density"/"intensity" and the amount of change thereof in each region based on the blood vessel pulsation cycle (step R5).

Next, noise removal is performed on the result obtained by fourier analysis or analysis of the adjustable coincidence ratio (step R6). Here, the deletion or the removal of the artifact (artifact) can be performed as described above. The operations of the steps R5 to R6 are performed 1 or more times to determine whether or not the operations are completed (step R7). Here, since the characteristic amount displayed on the display is mixed with the synthetic wave or other waves, there is a case where a frequency modulation image of a high-purity element, for example, a respiratory element, a blood flow element, or other elements cannot be displayed by extracting the frequency spectrum once. In this case, the feature amount displayed on the display may be used as a pixel value, and the analysis may be performed again for all or a part of the display. By this operation, it is possible to further acquire an image with high purity relating to the adjustability or consistency of the factor, for example, the respiratory factor or the blood flow factor. In this operation, the operator may manually recognize the image on the display while visually recognizing the image, or may automatically extract the spectrum from the output result and recalculate the distribution ratio. Even after the calculation, the noise removal processing, the space filling (interpolation) by the least square method, and the correction using the "density" of the surroundings may be performed according to the case.

In step R7, in the event of incompletion, the process proceeds to step R5, and in the event of completeness, the result obtained by fourier analysis or analysis of the adjustable coincidence rate is displayed on the display as a false-color image (step R8). In addition, black and white images may also be displayed. This can improve the accuracy of the data. In addition, a desired video may be obtained by correcting an image displayed on the display.

In the present embodiment, the desired frequency or frequency band is calculated by calculation, but if viewed as an actual image, the display is not necessarily limited to a good image. Therefore, the following method may be used.

(1) A method of presenting a plurality of frequency bands and selecting by a person;

(2) a method of prompting a plurality of frequencies and extracting a good image by using pattern recognition through an AI technique;

(3) the selection is made according to the tendency and morphology of HISTGRAM. That is, the value of the center portion of the "histogram" in the resulting signal tends to be high, and the value of the "histogram" fluctuates according to the motion, and therefore, the selection may be made according to the tendency and form of the histrgram.

When performing the analysis by 3D, the respiration volume, the cardiac output, and the central blood flow are measured by another device, and the respiration volume, the cardiac output, and the central blood flow in each block region can be calculated from the fourier analysis result, which is a relative value. That is, in the case of respiratory function analysis, the pulmonary ventilation amount can be estimated from the respiratory amount, in the case of pulmonary blood flow analysis, the pulmonary blood flow amount can be estimated from the cardiac (pulmonary blood vessel) output amount, and in the case of other blood flow amount analysis, the estimated blood flow amount (ratio) in the branch blood vessel drawn from the blood flow amount (ratio) on the central side can be estimated.

As described above, the acquired database can be determined with higher accuracy when all the data can be calculated, but it may take time to perform the computer analysis. Therefore, it is also possible to extract and calculate only an arbitrary number of sheets (for example, a specific stage). This can shorten the analysis time and remove irregular parts observed at the start of breathing. In addition, when the analysis result is displayed, an arbitrary range may be displayed. For example, by displaying a range from the transition of "expiration/inspiration" to the transition of "inspiration/expiration", so-called "infinite playback" is realized upon repeated playback, whereby diagnosis by a doctor can be facilitated.

As described above, according to the present embodiment, it is possible to evaluate an image of a human body using an X-ray imaging apparatus. If digital data is available, the rough calculation can be done well with existing facility equipment, leading to low import costs. For example, in an X-ray imaging apparatus using a flat panel detector, inspection of an object can be easily completed. In addition, pulmonary thromboembolism can be screened for pulmonary blood flow. For example, in an X-ray imaging apparatus using a flat panel detector, unnecessary examination can be eliminated by executing the diagnosis support program according to the present embodiment before CT is performed. In addition, because the examination is simple, patients with high urgency can be found at an early stage, and can be treated with priority. In the imaging method at the present time, although some problems are present in other modalities such as CT and MRI, if these problems are solved, precise diagnosis of each region is realized.

In addition, the method can be applied to screening for narrowing blood flow in various vessels, for example, in the neck, and also to blood flow evaluation and screening for large vessels. The lung respiration data is effective for a partial lung function test, and can be used as a lung function test. In addition, patients with COPD, emphysema, and the like can be identified. In addition, the method can also be applied to the grasping of preoperative and postoperative conditions. Furthermore, by performing fourier analysis on the cycle of the respiratory element and the blood flow cycle and removing the respiratory waveform and the blood flow waveform from the X-ray image of the abdomen, it is possible to observe changes in the remaining biological motion, such as intestinal obstruction.

Further, when the image acquired at the beginning is high-definition to some extent, since there are many pixels, it may take a long time to calculate. In this case, the image may be reduced to a certain number of pixels and then calculated. For example, by making the pixel of "4096 × 4096" actually "1024 × 1024" and then performing the calculation, the calculation time can be suppressed.

[ others ]

In addition, when an X-ray image is taken, a prediction algorithm such as an AR method (Autoregressive Moving average model) may be used. If the frequency of at least one of the breathing elements can be determined, the X-ray recording device can be controlled to adjust the irradiation interval of the X-rays according to the frequency. For example, when the frequency of the respiratory component is small (when the cycle is long), the number of times of X-ray imaging can be reduced. This can reduce the amount of exposure to the human body. In addition, when the frequency of respiratory factors such as tachypnea and tachypnea or cardiovascular pulsation is large (when the cycle is short), the irradiation frequency may be increased to perform optimal image generation.

Further, as a storage form of DICOM data, the image quality may be degraded if compressed, and therefore, it is preferable to store the DICOM data in a non-compressed form. Further, the calculation method may be changed according to the compression form of the data.

Description of the reference symbols

1 basic module

3 respiratory function analysis unit

5 pulmonary blood flow analysis unit

7 other blood flow analyzing section

9 Fourier analysis unit

10 waveform analysis section

11 visualization and digitization part

13 input interface

15 database

17 output interface

19 display.