阅读说明：本技术 具有喉部余液检测功能的便携式喉镜及余液检测方法 (Portable laryngoscope with throat residual liquid detection function and residual liquid detection method ) 是由 陈世雄 黄为民 朱明星 黄保发 方贤权 黄�俊 于 2020-06-12 设计创作，主要内容包括：本申请公开了具有喉部余液检测功能的便携式喉镜,其中,图像采集设备用于采集喉部图像,吸液装置用于将喉液吸入液体吸引口,图像处理识别装置分别与图像采集设备和吸液装置电连接,用于从喉部图像中识别出属于液体区域的至少一个像素点,依据液体区域内各像素点之间的相同图像特征将喉部图像中具有与该相同图像性质的像素点与至少一个像素点进行归并,以识别出喉部图像中的液体区域,然后依据该液体区域的面积控制吸液装置停止吸液。本实施例通过图像采集设备实时获取患者喉部图像并利用图像检测识别的方式判断图像中喉液区域的面积来判断患者喉部液体的余液量,实现了自动判断喉液是否吸完并据此对吸液装置进行自适应调节。(The application discloses portable laryngoscope with remaining liquid detection function of throat, wherein, image acquisition equipment is used for gathering the throat image, imbibition device is used for inhaling the liquid suction port with throat liquid, image processing recognition device is connected with image acquisition equipment and imbibition device electricity respectively, be used for discerning at least one pixel that belongs to the liquid region in the follow throat image, the same image characteristic between each pixel in the liquid region is with having in the throat image with this same image property with this pixel merge with at least one pixel, in order to discern the liquid region in the throat image, then according to the regional area control imbibition device of this liquid and stop the imbibition. According to the embodiment, the image acquisition equipment acquires the laryngeal image of the patient in real time and judges the area of the laryngeal area in the image by using an image detection and identification mode to judge the residual liquid amount of the laryngeal liquid of the patient, so that whether the laryngeal liquid is completely absorbed or not is automatically judged, and the liquid absorption device is adaptively adjusted according to the residual liquid amount.)

1. A portable laryngoscope with residual liquid laryngeal detection, comprising:

the image acquisition equipment is arranged near a liquid suction port at the tail end of the lens of the laryngoscope and is used for acquiring images of the larynx;

a liquid suction device for sucking the laryngeal liquid into the liquid suction port;

and the image processing and identifying device is respectively electrically connected with the image acquisition equipment and the liquid suction device and is used for identifying a liquid area from the throat image and controlling the liquid suction device to stop sucking liquid according to the area of the liquid area.

2. The portable laryngoscope according to claim 1, wherein the image processing and identifying means identifies at least one pixel point belonging to a liquid region from the laryngeal image, merges a pixel point in the laryngeal image having the same image properties as the at least one pixel point according to the same image characteristics between pixel points in the liquid region to identify the liquid region in the laryngeal image, and then controls the liquid suction means to stop sucking liquid according to the area of the liquid region.

3. The portable laryngoscope according to claim 2, wherein the image processing and identification means comprises:

the image area traversal module is used for carrying out window traversal on the throat image and calculating the area color mean value of each window area;

the initial pixel determining module is used for taking a coordinate point in a region corresponding to the region color mean value with the minimum difference value with the color reference value as a seed point and taking the seed point as an initial liquid region pixel set;

the liquid pixel judgment module is used for calculating the difference value between the color value of each adjacent pixel point of the area formed by the liquid area pixel set and the color mean value of the liquid area pixel set, and adding the adjacent pixel point corresponding to the difference value smaller than the color difference threshold value into the liquid area pixel set so as to update the liquid area pixel set;

the liquid area aggregation module is used for performing iteration of the liquid pixel judgment step on the updated liquid area pixel set until the difference values corresponding to all adjacent pixel points of the liquid area pixel set are not smaller than the color difference threshold value, and then taking the latest liquid area pixel set as a liquid area;

and the liquid suction control module is used for controlling the liquid suction device to stop sucking liquid when the area of the liquid area is lower than a preset area threshold value.

4. The portable laryngoscope according to claim 3, wherein the colour reference value comprises a first reference value corresponding to a first colour liquid and a second reference value corresponding to a second colour liquid; and the number of the first and second electrodes,

the starting pixel determining module obtains a first liquid area pixel set and a second liquid area pixel set respectively based on the first reference value and the second reference value, the liquid pixel judging module updates the first liquid area pixel set and the second liquid area pixel set respectively, the liquid area gathering module obtains a first liquid area corresponding to first color liquid and a second liquid area corresponding to second color liquid respectively, and the imbibition control module controls the imbibition device to stop imbibition when the area of the first liquid area is lower than the area threshold of the first color liquid and the area of the second liquid area is lower than the area threshold of the second color liquid.

5. The portable laryngoscope according to claim 3 or 4, wherein the image processing and identification device further comprises: the gray processing module is used for graying the acquired throat image to obtain a gray image; and the number of the first and second electrodes,

the throat image subjected to window traversal by the image area traversal module is the gray image, and the color value, the color mean value, the color reference value and the color difference threshold value adopt corresponding gray values.

6. The portable laryngoscope according to claim 3 or 4, wherein if there are a plurality of regions corresponding to the region color mean value having the smallest difference from the color reference value, the start pixel determination module selects one of the plurality of regions as the region for selecting the seed point, or calculates the color mean value of each adjacent region of each of the plurality of regions, and uses the region corresponding to the color mean value of the adjacent region having the smallest difference from the color reference value as the region for selecting the seed point.

7. The portable laryngoscope according to claim 3 or 4, wherein the liquid pixel decision module, when updating the liquid region pixel set, also calculates and updates a color mean of the liquid region pixel set.

8. The portable laryngoscope according to any one of claims 1 to 7, wherein the laryngoscope further comprises:

a laryngoscope blade assembly, one end of which is provided with a connecting shell;

the liquid storage cylinder is connected with the connecting shell and used for storing throat liquid sucked by the liquid suction device;

the handle is detachably connected with the liquid storage cylinder; wherein the content of the first and second substances,

the imbibition device includes imbibition hose and suction drive arrangement, imbibition hose one end set up in being close to near the end of laryngoscope piece subassembly, the other end stretches into inside the liquid storage cylinder, suction drive arrangement install in inside the liquid storage cylinder, be used for the liquid storage cylinder inner space produces the negative pressure, image acquisition equipment install near imbibition port of imbibition hose.

9. The portable laryngoscope as recited in claim 8, wherein the barrel of the fluid storage barrel is provided with a through hole, an H-shaped sealing ring is mounted on the through hole, and the other end of the fluid suction hose penetrates through a central hole of the H-shaped sealing ring and is in clearance fit with the central hole.

10. A portable laryngoscope implemented residual fluid detection method according to any one of claims 1 to 9, comprising:

identifying a liquid region from a laryngeal image acquired by an image acquisition device mounted on a laryngoscope and located near a liquid suction port at the distal end of the lens, and controlling a liquid suction device to stop sucking liquid according to the area of the liquid region; wherein the liquid suction device is used for sucking throat liquid into the liquid suction port.

Technical Field

The application relates to the technical field of medical instruments, in particular to a portable laryngoscope with a laryngeal residual liquid detection function and a residual liquid detection method.

Background

The laryngeal part of the pharynx, especially the hypopharynx and the larynx, is a common cavity of the respiratory tract, the digestive tract and the vocal tract, and is responsible for ventilation, phonation, swallowing initiation, swallowing airway protection reflex and the like. In the process of oral diagnosis and treatment, the laryngeal position is deep, the physiological structure is complex, direct peeping cannot be achieved, and devices such as a laryngoscope and the like are needed in laryngeal examination. In the fields of anesthesia, emergency treatment, intensive therapy, and the like, a catheter is generally inserted into a trachea from the oral cavity to ensure the passage of oxygen required for respiration, i.e., the opening of a respiratory tract, i.e., a tracheal intubation technique. The tracheal intubation technology is mainly applied to emergency rescue treatment such as operative anesthesia, cardiopulmonary resuscitation, prevention of aspiration and respiratory dysfunction, and the success rate of tracheal intubation is directly related to whether an operation can be performed or not. At present, the tracheal intubation is generally completed by means of a laryngoscope clinically.

A laryngoscope is a device used by clinicians during endotracheal intubation and assists in intubation by allowing the clinician to see the path of the endotracheal tube as it passes through the glottis towards the trachea. Since the endotracheal tube may be inserted into the trachea of a patient leading to the lungs rather than into the esophagus leading to the stomach of the patient during intubation, intubation of the endotracheal tube with the aid of a laryngoscope is required to improve the intubation success rate in order to prevent the endotracheal tube from being inserted mistakenly into the esophagus or damaging the glottis.

The laryngoscope is mainly composed of a handle and a laryngoscope blade, wherein the laryngoscope blade is used for shifting the epiglottis and lifting the tongue root to fully expose the glottis. During the cardiopulmonary resuscitation operation of neonate, often need clear up the liquid (blood, phlegm etc.) in the baby's oral cavity, then can carry out the intubate operation, this process can be accomplished with the help of the laryngoscope of taking the imbibition function, and the imbibition function is realized through setting up the imbibition hose in the laryngoscope.

However, the whole cardiopulmonary resuscitation operation needs to be accurately completed in a short time, otherwise, the tongue root and carotid chemoreceptor excitation is caused by the lifting force of the laryngoscope and the stimulation to the tracheal wall when the tracheal cannula is inserted, so that the sympathetic adrenaline system and the renin angiotensin system are activated, more catecholamine and angiotensin are secreted, the reactions of vasoconstriction, hypertension and the like are caused, and the vital signs of a patient are influenced. Therefore, the operation time of liquid cleaning (liquid suction) is high when the intubation is operated.

In addition, utilize the laryngoscope to carry out the process of imbibition when intubate operation at present, mainly carry out the manual control imbibition through the manual work, start the opportunity etc. of opening including the imbibition and be doctor manual operation, degree of automation is lower, leads to doctor's the operation degree of difficulty and whole intubate operation consuming time all to increase to manual operation also does not benefit to the accuracy nature of imbibition. Therefore, how to accelerate the liquid cleaning of the throat and realize the automation of the cleaning process is a problem to be solved.

Furthermore, the existing laryngoscope with the liquid suction function has more devices participating in the liquid suction process and longer device chain, so that the laryngoscope is bulkier and is not easy to operate, but the size of the device is determined by the function performed by the laryngoscope, so that the convenience and accuracy of operation can be influenced to a certain extent, the difficulty of operation of a doctor during intubation operation can be increased by the laryngoscope with larger device size, and the operation duration of liquid cleaning can be further increased.

Disclosure of Invention

Object of the application

Based on this, in order to satisfy the length of time requirement to throat imbibition operation, the intubate operation of accurate completion in the time as short as possible improves the automation and the intelligent degree of imbibition process simultaneously to further improve imbibition accuracy nature and reduce imbibition consuming time, and improve the lightweight degree and the portable degree of laryngoscope, do benefit to personnel's operation, the application discloses following technical scheme.

(II) technical scheme

In one aspect, the application discloses a portable laryngoscope with residual liquid laryngeal detection, comprising:

the image acquisition equipment is arranged near a liquid suction port at the tail end of the lens of the laryngoscope and is used for acquiring images of the larynx;

a liquid suction device for sucking the laryngeal liquid into the liquid suction port;

and the image processing and identifying device is respectively electrically connected with the image acquisition equipment and the liquid suction device and is used for identifying a liquid area from the throat image and controlling the liquid suction device to stop sucking liquid according to the area of the liquid area.

In a possible embodiment, the image processing and identifying device identifies at least one pixel point belonging to a liquid region from the throat image, merges the pixel point having the same image property as the throat image with the at least one pixel point according to the same image feature between the pixel points in the liquid region to identify the liquid region in the throat image, and then controls the liquid suction device to stop liquid suction according to the area of the liquid region.

In one possible embodiment, the image processing and recognition device includes:

the image area traversal module is used for carrying out window traversal on the throat image and calculating the area color mean value of each window area;

the initial pixel determining module is used for taking a coordinate point in a region corresponding to the region color mean value with the minimum difference value with the color reference value as a seed point and taking the seed point as an initial liquid region pixel set;

the liquid pixel judgment module is used for calculating the difference value between the color value of each adjacent pixel point of the area formed by the liquid area pixel set and the color mean value of the liquid area pixel set, and adding the adjacent pixel point corresponding to the difference value smaller than the color difference threshold value into the liquid area pixel set so as to update the liquid area pixel set;

the liquid area aggregation module is used for performing iteration of the liquid pixel judgment step on the updated liquid area pixel set until the difference values corresponding to all adjacent pixel points of the liquid area pixel set are not smaller than the color difference threshold value, and then taking the latest liquid area pixel set as a liquid area;

and the liquid suction control module is used for controlling the liquid suction device to stop sucking liquid when the area of the liquid area is lower than a preset area threshold value.

In a possible embodiment, the color reference values comprise a first reference value corresponding to a first color liquid and a second reference value corresponding to a second color liquid; and the number of the first and second electrodes,

the starting pixel determining module obtains a first liquid area pixel set and a second liquid area pixel set respectively based on the first reference value and the second reference value, the liquid pixel judging module updates the first liquid area pixel set and the second liquid area pixel set respectively, the liquid area gathering module obtains a first liquid area corresponding to first color liquid and a second liquid area corresponding to second color liquid respectively, and the imbibition control module controls the imbibition device to stop imbibition when the area of the first liquid area is lower than the area threshold of the first color liquid and the area of the second liquid area is lower than the area threshold of the second color liquid.

In a possible implementation, the image processing and recognizing apparatus further includes: the gray processing module is used for graying the acquired throat image to obtain a gray image; and the number of the first and second electrodes,

the throat image subjected to window traversal by the image area traversal module is the gray image, and the color value, the color mean value, the color reference value and the color difference threshold value adopt corresponding gray values.

In a possible implementation manner, if there are a plurality of regions corresponding to the region color mean value with the minimum difference value from the color reference value, the start pixel determining module selects one of the plurality of regions as a region of the selected seed point, or calculates a color mean value of each adjacent region of each of the plurality of regions, and uses the region corresponding to the adjacent region color mean value with the minimum difference value from the color reference value as the region of the selected seed point.

In a possible implementation manner, when the liquid pixel judgment module updates the liquid region pixel set, the liquid pixel judgment module further calculates and updates a color mean value of the liquid region pixel set.

In a possible embodiment, the pipetting control module further controls the pipetting device to initiate pipetting when the liquid area is not below the area threshold, and adjusts the pipetting power as a function of a difference between the liquid area and the area threshold.

In one possible embodiment, the laryngoscope further comprises:

a laryngoscope blade assembly, one end of which is provided with a connecting shell;

the liquid storage cylinder is connected with the connecting shell and used for storing throat liquid sucked by the liquid suction device;

the handle is detachably connected with the liquid storage cylinder; wherein the content of the first and second substances,

the imbibition device includes imbibition hose and suction drive arrangement, imbibition hose one end set up in being close to near the end of laryngoscope piece subassembly, the other end stretches into inside the liquid storage cylinder, suction drive arrangement install in inside the liquid storage cylinder, be used for the liquid storage cylinder inner space produces the negative pressure, image acquisition equipment install near imbibition port of imbibition hose.

In a possible embodiment, the barrel of the liquid storage barrel is provided with a through hole, an H-shaped sealing ring is arranged on the through hole, and the other end of the liquid suction hose penetrates through a central hole of the H-shaped sealing ring and is in clearance fit with the central hole.

In a possible implementation mode, one end of the liquid storage cylinder, which is connected with the handle, is provided with a connecting end cover, and the liquid storage cylinder is in threaded connection with the handle through the connecting end cover.

In one possible embodiment, the laryngoscope blade assembly is hinged to the reservoir.

In another aspect, the application discloses a method of residual laryngeal fluid detection for laryngoscopes, comprising:

identifying a liquid region from a laryngeal image acquired by an image acquisition device mounted on a laryngoscope and located near a liquid suction port at the distal end of the lens, and controlling a liquid suction device to stop sucking liquid according to the area of the liquid region; wherein the liquid suction device is used for sucking throat liquid into the liquid suction port.

In a possible embodiment, the identifying a liquid region from the throat image and controlling the liquid suction device to stop sucking liquid according to an area of the liquid region includes:

identifying at least one pixel point belonging to a liquid area from the throat image, merging the pixel points with the same image characteristics as the at least one pixel point in the throat image according to the image characteristics of the pixel points in the liquid area to identify the liquid area in the throat image, and then controlling the liquid suction device to stop liquid suction according to the area of the liquid area.

In a possible embodiment, the identifying at least one pixel point belonging to a liquid region from the throat image, merging the pixel points in the throat image having the same image characteristics as the at least one pixel point according to the image characteristics of the pixel points in the liquid region to identify the liquid region in the throat image, and then controlling the liquid suction device to stop sucking liquid according to the area of the liquid region includes:

identifying a liquid region from the throat image, and controlling the liquid suction device to stop sucking liquid according to the area of the liquid region, comprising:

performing window traversal on the throat image, and calculating a region color mean value of each window region;

taking a coordinate point in a region corresponding to the region color mean value with the minimum difference value with the color reference value as a seed point, and taking the seed point as an initial liquid region pixel set;

calculating a difference value between a color value of each adjacent pixel point of an area formed by the liquid area pixel set and a color mean value of the liquid area pixel set, and adding the adjacent pixel point corresponding to the difference value smaller than a color difference threshold value into the liquid area pixel set so as to update the liquid area pixel set;

performing iteration of the liquid pixel judgment step on the updated liquid region pixel set until the difference values corresponding to all adjacent pixel points of the liquid region pixel set are not less than the color difference threshold value, and then taking the latest liquid region pixel set as a liquid region;

and controlling the liquid suction device to stop sucking liquid when the area of the liquid area is lower than a preset area threshold value.

In a possible embodiment, the color reference values comprise a first reference value corresponding to a first color liquid and a second reference value corresponding to a second color liquid; and obtaining a first liquid area pixel set and a second liquid area pixel set based on the first reference value and the second reference value respectively, updating the first liquid area pixel set and the second liquid area pixel set respectively to obtain a first liquid area corresponding to the first color liquid and a second liquid area corresponding to the second color liquid respectively, and controlling the liquid suction device to stop sucking liquid when the area of the first liquid area is lower than the area threshold of the first color liquid and the area of the second liquid area is lower than the area threshold of the second color liquid.

In one possible embodiment, the method further comprises: firstly, graying the collected throat image to obtain a gray image; and the throat image subjected to window traversal is the gray image, and the color value, the color mean value, the color reference value and the color difference threshold value adopt corresponding gray values.

In a possible implementation manner, if there are a plurality of regions corresponding to the color mean value of the region with the minimum difference value from the color reference value, one of the regions is selected as a region for selecting a seed point, or the color mean value of each adjacent region of each of the regions is calculated, and the region corresponding to the color mean value of the adjacent region with the minimum difference value from the color reference value is used as the region for selecting a seed point.

In one possible embodiment, when updating the liquid region pixel set, a color mean of the liquid region pixel set is also calculated and updated.

In a possible embodiment, the pipetting device is further controlled to initiate pipetting when the liquid area is not below the area threshold, and the pipetting power is adjusted in dependence on the difference between the liquid area and the area threshold.

(III) advantageous effects

The application discloses portable laryngoscope and remaining liquid detection method with remaining liquid detection function of larynx has following beneficial effect:

the image acquisition equipment acquires the laryngeal image of the patient in real time and judges the area of the laryngeal region in the image by using an image detection and identification mode to judge the residual liquid amount of the laryngeal liquid of the patient, and then automatically controlling the liquid suction device to finish liquid suction when no residual liquid exists, realizing the automatic judgment of whether the throat liquid is sucked up and carrying out self-adaptive adjustment on the liquid suction device according to the judgment, replacing the operation of observing the residual liquid amount by naked eyes of a doctor and manually stopping the liquid suction device by automatic control, avoiding the possible misoperation during manual operation, improving the automation and the intelligent degree of liquid suction, reducing the operation difficulty of the doctor, saving the fragmentation time of manual operation, the time consumed in the liquid suction operation in the whole intubation operation process is reduced to some extent, the requirement of the intubation operation on the time of the liquid cleaning operation is met, the intubation success rate is improved, and a favorable starting point is provided for the follow-up examination and even operation of a patient by a doctor.

Through set up liquid (blood, phlegm etc.) clearance suction port on the laryngoscope piece, add the negative pressure and can attract the clearance to the liquid of oral cavity when using the laryngoscope, carry out the intubate operation after the glottis is clear after that, need not help handheld aspirator to reuse the laryngoscope after the clean up, practice thrift the whole time of intubate operation, have very important clinical meaning.

Adopt the external stock solution container that has great space of independent configuration to replace the liquid storage cylinder of built-in low capacity, the requirement to the liquid storage volume when the low capacity liquid storage cylinder can satisfy the imbibition of single or a small amount of times, and can also wash the liquid storage cylinder when wasing the imbibition device, avoided cross infection promptly, phenomenon such as propagation of bacterium and virus takes place, can not practice thrift the time of wasing external stock solution container alone again, the lightweight degree and the portable degree of laryngoscope have still been improved simultaneously, do benefit to personnel's operation.

Drawings

The embodiments described below with reference to the drawings are exemplary and intended to be used for explaining and illustrating the present application and should not be construed as limiting the scope of the present application.

Fig. 1 is a block diagram of a first embodiment of a portable laryngoscope with residual laryngeal liquid detection as disclosed herein.

Fig. 2 is a block diagram showing the construction of the second and third embodiments of the portable laryngoscope with residual liquid laryngeal detection disclosed in the present application.

Fig. 3 is a schematic three-dimensional view of a fourth embodiment of a portable laryngoscope with residual laryngeal liquid detection as disclosed herein.

Fig. 4 is a schematic cross-sectional view from the left perspective in fig. 3.

Fig. 5 is a schematic diagram of the three-dimensional structure of fig. 4.

Fig. 6 is a schematic three-dimensional structure of a foldable portable laryngoscope.

Fig. 7 is a schematic flow diagram of a first embodiment of a method of residual laryngeal fluid detection for laryngoscopes according to the present disclosure.

Fig. 8 is a schematic flow chart diagram of another embodiment of the first embodiment of the method for detecting residual laryngeal solution for laryngoscopes disclosed herein.

Fig. 9 is a schematic flow diagram of yet another embodiment of the first embodiment of the method for detecting residual laryngeal solution for laryngoscopes disclosed herein.

Detailed Description

In order to make the implementation objects, technical solutions and advantages of the present application clearer, the technical solutions in the embodiments of the present application will be described in more detail below with reference to the drawings in the embodiments of the present application.

A first embodiment of the portable laryngoscope with residual liquid laryngeal detection disclosed herein is described in detail below with reference to fig. 1, 3 and 4. As shown in fig. 1, the portable laryngoscope disclosed in this embodiment mainly includes: the device comprises an image acquisition device, a liquid suction device and an image processing and identifying device.

The image acquisition device is arranged near a liquid suction port at the tail end of the lens of the laryngoscope and is used for acquiring images of the larynx. In order to extend into the throat with one end of the laryngoscope lens, as shown in fig. 3 and 4, the image capturing device is a smaller device, such as a miniature camera 400, a pinhole camera, etc., which is sized to conveniently extend into the throat of a patient without causing trauma, and is specifically installed at the lower side of the laryngoscope lens 110 where the liquid suction hose 300 penetrates. The camera can be provided with an illumination light source, and can also illuminate through an external light source.

The liquid suction device is used for sucking throat liquid into the liquid suction port. The pipetting device generally comprises a member for providing a pipetting channel and a member for providing a driving force for a suction force, as shown in fig. 3 and 4, using a pipetting hose 300 having one end extended into a closed space, and a negative pressure suction motor 210 or a negative pressure ball or other device or member capable of providing a suction force for a liquid (sputum, etc.) to be aspirated, which is installed in the closed space. Wherein, the liquid suction port is a liquid suction port of the liquid suction hose 300, the negative pressure suction motor can suck the liquid into the hose by generating the negative pressure, and the negative pressure suction motor and the hose can be respectively and selectively arranged inside the structure of the laryngoscope or outside the structure of the laryngoscope.

The image processing and identifying device is electrically connected with the image acquisition equipment and the liquid suction device respectively. The image processing and identifying device can be directly arranged on the structural body of the portable laryngoscope, namely integrally connected with the portable laryngoscope, for example, arranged at the tail part of the handle, at the moment, a micro or capsule type high integration device is required to avoid increasing the volume and the operation difficulty of the laryngoscope, under the condition that the image processing and identifying device is directly arranged on the handle of the laryngoscope instead of being in remote communication connection with the image acquisition device and the liquid suction device, the size of the image processing and identifying device is subject to no remarkable increase of the operation difficulty of a doctor, and particularly, a TMPV7502XBG image identification processor with the size of 11mm multiplied by 11mm of Toshiba company and the like can be adopted. The image processing and identifying device can also be connected with the portable laryngoscope in a split way, and the image processing and identifying device can be communicated with the image acquisition equipment and the liquid suction device on the laryngoscope in a light ray connection or wireless communication mode.

The image processing and identifying device is used for identifying a liquid region from the throat image and controlling the liquid suction device to stop liquid suction according to the area of the liquid region.

The method has the advantages that the image acquisition equipment acquires the image of the throat of the patient in real time, the area of the throat liquid area in the image is judged by utilizing the image detection and identification mode to judge the residual liquid amount of the throat liquid of the patient, then the liquid suction device is automatically controlled to finish liquid suction when no residual liquid exists, whether the throat liquid is sucked up or not is automatically judged, self-adaptive adjustment is carried out on the liquid suction device according to the automatic judgment, the automation and the intelligent degree of liquid suction are improved, the operation difficulty of a doctor is reduced, the fragmentation time of manual operation is saved, the time consumed in the whole intubation operation process on the liquid suction operation is reduced, the requirement of intubation operation on the time of liquid cleaning operation is met, the intubation success rate is improved, and a favorable starting point is provided for the follow-up examination and even operation.

Specifically, the image processing and identifying device is used for identifying at least one pixel point belonging to the liquid region from the throat image, merging the pixel point having the same image property with the at least one pixel point in the throat image according to the same image characteristics among the pixel points in the liquid region to identify the liquid region in the throat image, and then controlling the liquid absorbing device to stop liquid absorbing according to the area of the liquid region.

When the doctor need utilize the laryngoscope to inhale the larynx liquid of patient's larynx when carrying out the intubate operation, put into patient's larynx with the lens end of laryngoscope, image acquisition equipment begins to gather the larynx image in real time to upload the larynx image to image processing recognition device, contain the larynx liquid region in the larynx image this moment, also be exactly the liquid region. After the image processing and identifying device receives the laryngeal image, at least one pixel point which is bound to belong to the laryngeal area is identified as a starting point or a growing point, a region growing algorithm based on a threshold value is utilized, the single pixel point which belongs to the laryngeal area or an adjacent region which belongs to the laryngeal area starts, the pixel points are gathered according to the similar property of the pixels in the laryngeal area, the points of the adjacent region are merged into the current region, the laryngeal area in the laryngeal image is obtained, and the liquid in the laryngoscope image and the background image are identified and segmented. After the identification and the segmentation are finished, judging the area (pixel number) of the throat liquid area, if the area is larger, indicating that the residual liquid amount of the throat is larger, keeping the liquid suction device started to continuously suck liquid, and continuously carrying out identification and segmentation on the liquid area on the throat image at the next moment uploaded by the image acquisition equipment; if the area becomes very small or almost no, the residual liquid amount in the throat is very small or the residual liquid is completely sucked, and at the moment, the image processing and identifying device sends a signal instruction to the liquid suction device to control the liquid suction device to close so as to finish liquid suction.

The embodiment obtains the throat image of the patient in real time through the image acquisition equipment and judges the area of the throat liquid area in the image by using an image detection and identification mode to judge the residual liquid amount of the throat liquid of the patient, thereby automatically controlling the liquid suction device to finish liquid suction when no residual liquid exists, realizing the automatic judgment of whether the throat liquid is sucked up or not and carrying out self-adaptive adjustment on the liquid suction device according to the automatic judgment, avoiding the potential problem caused by too much and too little suction, replacing the operation of observing the residual liquid amount by naked eyes of a doctor and manually stopping the liquid suction device by automatic control, avoiding the misoperation possibly generated in the manual operation, improving the automation and the intelligent degree of liquid suction, reducing the operation difficulty of the doctor, saving the fragmentation time of the manual operation, reducing the time consumed in the whole intubation operation process on the liquid suction operation, and meeting the requirement of the intubation operation on the liquid cleaning operation time, the intubation success rate is improved, and a favorable starting point is provided for the subsequent examination and even operation of a patient by a doctor.

A second embodiment of a portable laryngoscope with residual laryngeal liquid detection disclosed herein is described in detail below with reference to fig. 2 to 4. As shown in fig. 2, the portable laryngoscope disclosed in this embodiment mainly includes: the device comprises an image acquisition device, a liquid suction device and an image processing and identifying device.

The image acquisition equipment is arranged near a liquid suction port of the laryngoscope, which is positioned at the tail end of the lens, and is used for acquiring images of the larynx, and the liquid suction device is used for sucking the laryngeal liquid into the liquid suction port. The installation positions, functions and the like of the image acquisition device, the liquid suction device and other components of the embodiment can refer to the structural arrangement described in the first embodiment of the portable laryngoscope, and are not described in detail herein. As shown in fig. 3 and 4, the image capturing device may employ a camera 400, and may be specifically mounted on the lower side of the laryngoscope blade 110 where the liquid suction hose 300 penetrates. The suction means may employ a suction hose 300 having one end extended into the closed space and a negative pressure suction motor 210 installed in the closed space.

The main difference between this embodiment and the first embodiment of the portable laryngoscope is that the image processing and identifying device in this embodiment comprises an image area traversing module, a starting pixel determining module, a liquid pixel judging module, a liquid area gathering module and a liquid suction control module.

The image area traversal module is used for performing window traversal on the throat image acquired by the image acquisition equipment and calculating the area color mean value of each window area.

After the image processing and identifying device receives the laryngeal image uploaded by the image acquisition equipment, an image area traversing module of the image processing and identifying device can perform window traversal on the laryngeal image by adopting a preset rectangular window, and the color average value of all pixel points in the rectangular window is calculated. For example, if the size of the rectangular window is 4 × 4 and the size of the throat image is 800 × 800, 797 × 797 pixel point matrices are obtained, and the color mean of 16 pixel points in each pixel point matrix is calculated to obtain 797 × 797 area color mean. The size of the rectangular window may be preset and unchangeable, or may be adaptively adjusted according to the size of the traversed throat image.

The starting pixel determining module is used for taking a coordinate point in the window area corresponding to the area color mean value with the minimum difference value with the color reference value as a seed point, and taking the seed point as an initial liquid area pixel set.

The color reference value is a color value set according to the color of the throat liquid, for example, the color of the throat liquid is usually white, and the brightness of the color of the liquid area is off white, so the color reference value can be set as an off white color value, and the pixel value of the color is large; if the color of the throat fluid contains yellow, the color reference value may be set to a color value of yellow, and the pixel value of the color is smaller than the pixel value of white. This embodiment will be described in detail by taking, as an example, a case where the color of the throat liquid is white.

After the image area traversal module obtains 797 × 797 area color mean values, the initial pixel determination module performs difference calculation on each area color mean value and a color reference value used as a reference to obtain 797 × 797 difference values, selects a minimum difference value min from the difference values, and determines an area color mean value Amin corresponding to the minimum difference value min, where the area color mean value Amin is an area color mean value closest to a color reference value Cwhite (white). And then determining a 4 × 4 pixel matrix Rseed corresponding to the Amin from 797 × 797 pixel matrixes according to the area color mean value Amin, wherein coordinate points (namely pixel points) in the pixel matrix Rseed are seed points for area growth and aggregation later.

The seed point can be selected by selecting a single pixel point from the pixel point matrix Rseed as the seed point, or selecting a plurality of pixel points from the seed point, or even all the pixel points as the seed points. If a single pixel point is selected as a seed point, the rule of selection may be to take a coordinate point of an intermediate value in the matrix Rseed as the seed point.

When the residual liquid amount of the throat image is judged every time, the image processing and identifying device establishes a liquid area pixel set Mwhite for gathering pixel points judged to belong to the liquid area, since the determined pixel matrix Rseed is closest to the color reference value of white, and the image content displayed in the laryngeal image generally includes up to a laryngeal organ tissue image, a liquid image and the like, wherein only the liquid image is white, and the color of other images such as the laryngeal organ tissue image is red, which is different from white and is darker, therefore, the pixel matrix Rseed must contain the pixels belonging to the liquid region, even all the pixels contained in the pixel matrix Rseed belong to the liquid region, therefore, the seed point selected from the matrix Rseed must belong to the liquid region, i.e. the first added pixel point in the pixel set Mwhite.

The liquid pixel judgment module is used for calculating the difference value between the color value of each adjacent pixel point of the area formed by the liquid area pixel set and the color mean value of the liquid area pixel set, and adding the adjacent pixel point corresponding to the difference value smaller than the color difference threshold value into the liquid area pixel set so as to update the liquid area pixel set.

The liquid pixel judgment module mainly has the function of a liquid pixel judgment step, namely guiding the seed points to grow and gather according to the growth and gathering rule. Because the liquid pixel judgment module carries out the first guidance of seed point growth and aggregation on a certain throat image, only seed points (the first batch of pixel points added with Mwhite) exist in the pixel set Mwhite at the moment, and the difference calculation and comparison for the first time actually carries out difference calculation on the color values of adjacent pixel points and the color mean value of the seed points.

Assuming that the seed point determined by the initial pixel determination module is a pixel region px _ area _1 containing four pixel points and having coordinates from (300, 500) to (301, 501), the region px _ area _1 formed by the pixel set of the liquid region is an initial liquid region and contains four pixel points, color values of 12 adjacent pixel points around px _ area _1 are determined, and difference calculation is performed between the color values and the color mean value of px _ area _1 region. And the adjacent pixel points refer to pixel points which have a difference of 1 unit with the target pixel points on the coordinates of the horizontal axis and/or the vertical axis. If the 12 pixel points contain pixel points which also belong to the liquid area image, the color values of the pixel points are also white, so that the difference value between the color values of the pixel points and the color mean value of the Mwhite is very small, and if the color difference is smaller than a set color difference threshold value, the pixel points belong to the liquid area, so that the pixel points are added into the pixel set Mwhite to obtain a new pixel set Mwhite. Assuming that 10 pixels in the three rows of the 12 pixels, i.e., the upper row, the lower row and the left row, all meet the requirement of the color difference threshold, the updated pixel set Mwhite will include 14 pixels.

It is understood that the color difference threshold may be preset and unchangeable, or may be adaptively adjusted according to situations such as different background states caused by different specific positions of the throat of a specific patient, and detected ambient brightness.

It should be noted that, if the seed point determined by the initial pixel determination module is a single pixel point, the pixel region px _ area _1 (initial liquid region) is a single pixel point, the color mean of px _ area _1 is the color value of the single pixel point, and the adjacent pixel points of px _ area _1 are the peripheral eight pixel points in the nine-square grid.

And the liquid area aggregation module is used for controlling the liquid pixel judgment module to carry out iteration of the liquid pixel judgment step on the updated liquid area pixel set until the corresponding difference values of all adjacent pixel points of the liquid area pixel set are not less than the color difference threshold value, and then taking the latest liquid area pixel set as a liquid area.

The liquid pixel judgment step executed by the liquid pixel judgment module is single, but if a complete liquid image in the throat image is obtained, repeated iterations are required, that is, the updated liquid region pixel set output by the liquid pixel judgment module is repeatedly substituted into the liquid pixel judgment module as the input initial liquid region pixel set, the above steps are repeated until the liquid pixel judgment module calculates the difference value of the current adjacent pixel points, finding that all the adjacent pixel points do not meet the requirement of the color difference threshold value, which indicates that all the adjacent pixel points do not belong to the liquid area at the moment, and the existing pixel points in the current liquid region pixel set are all the pixel points of the liquid region in the larynx liquid image, the fluid region has thus been identified and segmented from the laryngeal image at this point, and the iterative process can be stopped.

It should be noted that, when the liquid pixel judgment module is controlled to perform the difference calculation of the adjacent pixel points, only the adjacent pixel points which are not calculated are selected to perform the difference calculation, and the adjacent pixel points which are already calculated are not repeatedly calculated, so that the calculation time is saved, the judgment speed is increased, and the update speed of the liquid region pixel set is increased. For example, taking the above-mentioned adjacent pixel points as eight peripheral pixel points in a nine-square grid as an example, when the pixel point (200 ) at the upper left corner of the eight pixel points is taken as a target point, the adjacent pixel point of the target point is determined and the difference value calculation is performed, because the adjacent pixel points at the lower right side of the target point have been subjected to the difference value calculation, the difference value calculation is not performed at this time, and because the adjacent pixel point (201, 199) at the upper right side of the target point is also the adjacent pixel point of another target point (201, 200) at the adjacent right side of the target point, the difference value calculation is performed only once for the adjacent pixel point (201, 199) in the liquid pixel judgment process of this time, the difference value calculation for the adjacent pixel point is not required in the subsequent iteration process, specifically, the coordinate point subjected to the difference value calculation can be marked in a pixel coordinate, avoiding duplicate calculations.

And the liquid suction control module is used for controlling the liquid suction device to stop sucking liquid when the area of the liquid area is lower than a preset area threshold value.

After the liquid region aggregation module identifies and segments the liquid region from the throat liquid image, the liquid absorption control module obtains the area of the liquid region, which is equivalent to calculating the number of pixel points of the liquid region. The image processing and recognizing device is preset with an area threshold value, and the area threshold value is usually set to be lower. The liquid area gathering module can compare the number of pixel points contained in the liquid area with an area threshold value, if the number of pixel points is lower than the area threshold value, the remaining liquid amount of the throat part is reduced to a state of complete liquid suction, at the moment, liquid suction can be finished as soon as possible, a control signal is sent to the liquid suction device to control the liquid suction device to stop running so as to finish liquid suction, so that the requirement of the intubation operation on the liquid cleaning operation duration is met, time is won for the subsequent steps of the intubation operation, and a good starting point is created.

The embodiment discloses a scene for cleaning liquid of throat liquid with one color, a region growing algorithm based on a threshold value is adopted, the selection of seed points is interfered through image region traversal and mean difference value calculation, so that a better growing starting point is selected, pixels around the growing starting point are subjected to gradual color difference value judgment in an outward radiation mode, pixel points belonging to the same liquid are gathered to a liquid region pixel set, the liquid region grows from point to surface, the liquid region is finally obtained, and accordingly, the control of a liquid suction device is realized, the characteristic of obvious color difference between the liquid region and the background color of the throat is fully utilized, the identification and segmentation precision of the liquid region is improved, and the calculation result is accurate and quick.

In one embodiment, the image processing and recognition apparatus further comprises: and the gray processing module is used for graying the acquired throat image to obtain a gray image. And the throat image traversed by the window of the image area traversing module is a gray image, and the color value, the color mean value, the color reference value and the color difference threshold value adopt corresponding gray values.

Since the color of the throat liquid in this embodiment is white, the brightness of the color reflected in the image is white, the pixel value is high before the graying, and the gray value is high after the graying. The background color after graying, for example, the gray value of the laryngeal organ tissue is smaller, and graying can be beneficial to more accurately and more quickly distinguishing the laryngeal liquid with larger color brightness from the non-laryngeal liquid part with smaller color brightness, so that graying is firstly carried out, and then the subsequent step operation is carried out based on the image after graying.

Specifically, the gray processing module receives the throat image uploaded by the image acquisition device, and performs gray processing to obtain the throat gray image. And the image area traversing module performs window traversal on the throat gray level image and calculates the area gray level mean value of 797 multiplied by 797 window areas. The starting pixel determination module performs difference calculation on the gray level mean value of each region and a gray level reference value, the gray level reference value may be 255, for example, to obtain 797 × 797 difference values, determine a minimum difference value, and further determine a region gray level mean value corresponding to the minimum difference value, then determine a region corresponding to the region gray level mean value, select a region formed by one or more coordinate points from the region as a seed point, and use the seed point as an initial liquid region pixel set. The liquid pixel judgment module determines adjacent pixel points of the pixel set, calculates the difference value between the gray value of each adjacent pixel point and the gray average value of the pixel set, and adds the adjacent pixel points corresponding to the difference value smaller than the gray difference threshold value into the pixel set to realize the initial growth and aggregation of the seed points of the pixel set, wherein the gray difference threshold value can be selected within the range of [10, 50 ]. The liquid region aggregation module controls the liquid pixel judgment module to iterate the liquid pixel judgment step so as to realize continuous growth and aggregation of the pixel set until the difference values corresponding to all adjacent pixel points of the pixel set are not smaller than the gray difference threshold value, namely, the pixel points meet the region of non-larynx liquid in each growth direction, and at the moment, a complete liquid region is obtained.

It can be understood that, after performing the gray scale processing, the gray scale processing module may further perform normalization processing and gaussian filtering denoising processing on the obtained gray scale image to obtain a gray scale image with a standard size, for example, if the size of the obtained gray scale image is 400 × 400, the number of pixel point matrixes obtained under a 4 × 4 rectangular window is 397 × 397, so that the recognition and segmentation speed is increased on the premise of ensuring that the image recognition and segmentation accuracy is not affected.

In an embodiment, if there are a plurality of regions corresponding to the color mean of the region with the minimum difference from the color reference value, the starting pixel determining module selects one of the plurality of regions as the region for selecting the seed point, or calculates the color mean of each adjacent region of each of the plurality of regions, and uses the region corresponding to the color mean of the adjacent region with the minimum difference from the color reference value as the region for selecting the seed point.

Before the starting pixel determining module selects the seed point, the region corresponding to the region color mean value with the minimum difference value with the color reference value is determined, namely, the window region is selected. However, there may be a case where the area color mean or the area gray scale mean of the plurality of window regions is completely the same, or even a case where the color values or the gray values of the pixels included in the plurality of window regions are the same. To cope with this, the start pixel determination module has the following two ways of determining the region from which the seed point is selected.

First, a random selection is made from the plurality of window regions, and a seed point is selected from the selected region, thereby saving calculation time.

Secondly, expanding the selection criteria, assuming that the area mean values of 2 areas a1 and a2 are the same, obtaining the area mean values of eight adjacent window areas around a1, and then calculating the mean values of the eight area mean values to obtain the neighborhood mean value, and a2 is the same. And then, difference calculation is carried out on the neighborhood mean values of A1 and A2 and the reference value, and if the difference between the neighborhood mean value of A1 and the reference value is minimum, A1 is selected as a region for selecting the seed point. Therefore, the region for selecting the seed points is close to the center of the liquid region as much as possible, the distance between the seed points and the outermost edge of the liquid region is shortened, the iteration times of the subsequent liquid region during gathering are reduced, and the operation time is saved.

In one embodiment, when the liquid pixel judgment module updates the pixel set of the liquid region, the color mean of the pixel set is calculated and updated in addition to updating the pixel points included in the pixel set. If the color mean value of the pixel set is not updated, the color mean value of the pixel set adopted by the liquid pixel judgment module when the liquid pixel judgment is carried out every time is the color mean value of the seed point, and therefore when the throat light condition is poor and the color of the organ of Beijing is shown due to the thin liquid thickness, and the color value of the liquid is reduced, misjudgment is easy to generate.

In one embodiment, the pipetting control module further controls the pipetting device to initiate pipetting when the liquid region area is not below an area threshold, and adjusts the pipetting power based on a difference between the liquid region area and the area threshold.

The liquid suction control module closes the liquid suction device when the area of the liquid area is lower than an area threshold value, and can automatically judge whether the area of the liquid area is not lower than a preset area threshold value or not after a laryngoscope is placed in the throat of a patient, if the area of the liquid area is not lower than the area threshold value, liquid suction is required, and if the liquid suction device is not started at the moment, the liquid suction control module controls the liquid suction device to be started to suck liquid.

After the imbibition is started, the imbibition control module can calculate the difference between the area of the liquid region and the area threshold in real time, if the difference is larger, the residual liquid amount is larger, and at the moment, whether the imbibition power needs to be improved or not can be judged according to the current power of the imbibition device so as to accelerate the imbibition speed; if the difference value is smaller, it is shown that although throat liquid still exists, the liquid amount is smaller, at the moment, whether the liquid suction power needs to be reduced or not can be judged according to the current power of the liquid suction device, the damage to the throat part possibly caused by the suction in a high-power state is avoided, the liquid suction power gently rises and falls along with the size of the residual liquid amount through devices such as a micro frequency converter, the liquid suction safety is improved, and under the condition that the micro frequency converter is directly arranged on a laryngoscope handle instead of being in remote communication connection with other components of the liquid suction device, the size of the micro frequency converter is subject to the condition that the operation difficulty of a doctor is.

A third embodiment of the portable laryngoscope with residual liquid laryngeal detection disclosed herein is described in detail below with reference to fig. 2 to 4. As shown in fig. 2, the portable laryngoscope disclosed in this embodiment mainly includes: the device comprises an image acquisition device, a liquid suction device and an image processing and identifying device.

The image acquisition equipment is arranged near a liquid suction port of the laryngoscope, which is positioned at the tail end of the lens, and is used for acquiring images of the larynx, and the liquid suction device is used for sucking the laryngeal liquid into the liquid suction port. The installation positions, functions and the like of the image acquisition device, the liquid suction device and other components of the embodiment can refer to the structural arrangement described in the second embodiment of the portable laryngoscope, and are not described in detail herein. As shown in fig. 3 and 4, the image capturing device may employ a camera 400, and may be specifically mounted on the lower side of the laryngoscope blade 110 where the liquid suction hose 300 penetrates. The suction means may employ a suction hose 300 having one end extended into the closed space and a negative pressure suction motor 210 installed in the closed space.

The main difference between this embodiment and the second embodiment of the portable laryngoscope is that the color of the laryngeal fluid in this embodiment includes two colors, for example, yellow laryngeal fluid, which may appear in addition to white laryngeal fluid.

The image processing and identifying device in the embodiment comprises an image area traversing module, a starting pixel determining module, a liquid pixel judging module, a liquid area gathering module and a liquid absorption control module. Wherein the color reference values include a first reference value corresponding to the first color liquid and a second reference value corresponding to the second color liquid. And the initial pixel determining module obtains a first liquid area pixel set and a second liquid area pixel set respectively based on a first reference value and a second reference value, the liquid pixel judging module updates the first liquid area pixel set and the second liquid area pixel set respectively, and the liquid area gathering module obtains a first liquid area corresponding to the first color liquid and a second liquid area corresponding to the second color liquid respectively.

Specifically, the gray processing module receives the throat image uploaded by the image acquisition device, and performs gray processing to obtain the throat gray image. And the image area traversing module performs window traversal on the throat gray level image and calculates the area gray level mean value of 797 multiplied by 797 window areas.

Then, the initial pixel determination module performs difference calculation on the gray level mean value of each region and a first reference value representing white liquid to obtain a first group of 797 × 797 difference values, determines a first minimum difference value, further determines a region gray level mean value corresponding to the first minimum difference value, then determines a region corresponding to the region gray level mean value, selects a region formed by one or more coordinate points from the region as a first seed point corresponding to the white liquid, and determines a first batch of white liquid pixel points by taking the first seed point as an initial first liquid region pixel set, namely a white liquid region pixel set; the initial pixel determination module also determines a second seed point corresponding to the yellow liquid in the same way, and the second seed point is used as an initial second liquid area pixel set, namely the yellow liquid area pixel set, so that the determination of the first batch of yellow liquid pixel points is realized.

Then, the liquid pixel judgment module determines adjacent pixel points of a first liquid area pixel set, calculates the difference value between the gray value of each adjacent pixel point and the gray average value of the first liquid area pixel set, and adds the adjacent pixel points corresponding to the difference value smaller than the gray difference threshold value into the first liquid area pixel set to realize the primary growth and aggregation of the white liquid area; the liquid pixel judgment module also carries out primary growth and aggregation on the second liquid region pixel set in the same way.

And then, the liquid area aggregation module controls the liquid pixel judgment module to respectively iterate the liquid pixel judgment steps according to the two liquid area pixel sets so as to realize the continuous growth and aggregation of the first liquid area pixel set and the second liquid area pixel set until the two pixel sets grow to the point that the corresponding difference value of all adjacent pixel points of the respective pixel sets is not less than the gray level difference threshold value of the corresponding color, and at the moment, a complete white liquid area and a complete yellow liquid area are obtained. The sizes of the liquid areas of the two colors may be different, so the corresponding iteration times of the two colors may also be different, but the two colors are required to complete respective iteration processes and obtain a complete liquid area, so that the area judgment can be performed later.

It should be noted that, if the laryngeal image is not subjected to grayscale processing before window traversal, the image region traversal module directly performs window traversal on the colored laryngeal image, and then the grayscale values are color values, and the grayscale to the determination also becomes color value determination. If the number of the colors is more than two, the image processing and recognizing device can recognize and divide the liquid areas of the plurality of colors in parallel, and the implementation is the same as that of the yellow liquid added in the present embodiment.

It can be understood that, if the liquid regions are identified and segmented for two or more colors, in the process of respective growth and aggregation of the liquid region pixel sets of different colors, respective pixel points that have been confirmed to belong to the color may meet, that is, boundary edges of the regions are in contact, for example, a pixel point that belongs to white liquid with coordinates (400, 500) and a pixel point that belongs to yellow liquid with coordinates (401, 500) meet, and at this time, expansion and detection in the outward direction are not performed at the meeting point, because an adjacent pixel point in the direction has been confirmed to be liquid of other colors in the identification process of other colors.

In addition, if only the liquid region of one color of the throat is detected, only a series of corresponding parameters and steps related to the color, such as a reference value and a threshold value of the color to be detected, need to be selected, and the implementation mode at this time is the same as the case of performing recognition and segmentation only on white in the second embodiment.

Unlike the first embodiment, the pipetting control module controls the pipetting device to stop pipetting when the area of the first liquid region is below the area threshold for the liquid of the first color and the area of the second liquid region is below the area threshold for the liquid of the second color.

That is to say, only when the residual liquid amount of white liquid and yellow liquid is extremely low, the imbibition process can be ended, and under the condition that any color of liquid is still residual, imbibition needs to be continued, and the next-time laryngeal image uploaded by the image acquisition equipment is identified and segmented, so as to ensure complete imbibition and complete imbibition of laryngeal liquid.

The embodiment discloses a scene for simultaneously cleaning liquid from throat liquids with various colors, and the scene adopts a parallel mode to respectively identify and segment the liquid with various colors, also fully utilizes the characteristic of obvious color difference between a liquid area and the background color of the throat, improves the identification and segmentation precision of the liquid area, and has accurate and quick calculation result.

A fourth embodiment of a portable laryngoscope with residual liquid laryngeal detection as disclosed herein is described in detail below with reference to fig. 3-6. As shown in fig. 3, the portable laryngoscope disclosed in this embodiment mainly includes: the image acquisition equipment, the liquid suction device and the image processing and identifying device further comprise a laryngoscope lens assembly 100, a liquid storage cylinder 200 and a handle 500, and the liquid suction device comprises a liquid suction hose 300 and a suction driving device. The image processing and identifying device can adopt the image processing and identifying device described in any one of the first embodiment to the third embodiment of the portable laryngoscope.

The main difference between this embodiment and the three previous embodiments is the arrangement of the mechanical mechanism parts of the laryngoscope. As shown in fig. 3, a connection shell 120 is disposed at one end of the laryngoscope blade assembly 100, the laryngoscope blade assembly 100 further comprises a laryngoscope blade 110, and the laryngoscope blade 110 is connected with the connection shell 120. The outside of laryngoscope piece 110 can be equipped with the silica gel cover (not shown in the figure), prevents that the rigidity direct contact vocal cords of laryngoscope piece and then harms the vocal cord, realizes the soft contact of laryngoscope piece 110 and vocal cord, prevents effectively that the vocal cord from receiving the damage, has realized the buffer protection to the vocal cord.

As shown in fig. 4, the liquid storage barrel 200 is connected to the connection housing 120, and the liquid storage barrel 200 is mainly used for storing throat liquid inhaled by the liquid suction device. The extending direction (axial direction) of the reservoir 200 and the extending direction (longitudinal direction) of the laryngoscope blade 110 may be set to about 90 degrees. Realize from stock solution function through liquid storage barrel 200, need not to go liquid discharge to other external containers in, increased the portability of laryngoscope, reduced the operation degree of difficulty. Meanwhile, because the liquid suction device of the laryngoscope needs to be cleaned frequently for ensuring sanitation, and the throat liquid amount sucked by a single time is usually less and can only be 2-3ml, the liquid storage cylinder 200 with small capacity internally arranged is adopted to replace an externally arranged liquid storage container with a larger space which is arranged independently, the liquid storage cylinder 200 with small capacity can meet the requirement on the liquid storage amount during liquid suction of a single time or a small number of times, and the liquid storage cylinder 200 can be cleaned simultaneously when the liquid suction device is cleaned, so that the phenomena of cross infection, propagation of bacteria and viruses and the like are avoided, the time for cleaning the externally arranged liquid storage container independently is not saved, the light weight degree and the portability degree of the laryngoscope are improved, and the operation by personnel is facilitated.

As shown in fig. 4, the handle 500 is detachably connected with the liquid storage barrel 200, so that when the liquid storage amount of the liquid storage barrel 200 is large or the liquid storage barrel is full, the liquid storage barrel 200 can be detached and the stored liquid can be discharged, and the portable use and the reusable use of the liquid suction function of the laryngoscope can be realized. Specifically, as shown in fig. 5, a plurality of first external teeth 240 may be circumferentially and uniformly distributed on the exposed wall surface of the liquid storage cylinder 200 for increasing the friction force of screwing, so as to facilitate screwing-type detachment. The whole liquid storage cylinder 200 can be made of transparent materials, or the middle part of the liquid storage cylinder can be made of transparent materials, or a transparent window made of transparent materials can be arranged in the middle part of the liquid storage cylinder, so that a doctor can conveniently check the amount of liquid storage in the liquid storage cylinder 200, and the liquid storage cylinder 200 can be cleaned in time.

One end of a suction hose 300 is disposed near the distal end of the laryngoscope blade assembly 100 and extends out of the laryngoscope blade 110 for extending into the patient's throat to provide a suction channel. The other end of the suction hose 300 extends into the liquid storage barrel 200, and the suction driving device is installed inside the liquid storage barrel 200 and used for generating negative pressure in the inner space of the liquid storage barrel 200, so that the suction hose 300 can suck liquid in. The suction driving means may specifically employ a negative pressure suction motor 210 having a small volume and capable of being placed in the liquid cartridge 200. The negative pressure suction motor 210 may not directly contact the suction hose 300, and may transmit the negative pressure into the suction hose 300 only by the airtight space inside the liquid storage barrel 200, so that the suction hose 300 generates suction force to further suck the throat fluid.

In addition, in order to realize that the inner space of the liquid storage cylinder 200 is not only a closed space, but also can conveniently open the liquid storage cylinder 200 to clean and discharge the liquid storage inside, one end cover of the liquid storage cylinder 200 can be detachably arranged, or the end covers at the two ends are detachably arranged, or the liquid storage cylinder 200 is formed by sealing, splicing and installing a plurality of components.

The image acquisition device is installed near the liquid suction port of the liquid suction hose 300 and is used for acquiring images at the liquid suction port of the liquid suction hose 300 and uploading the images to the image processing and identifying device so as to control the operation of the suction driving device. The image acquisition device can adopt the camera 400 shown in fig. 3, the camera 400 can be in communication connection with the image processing and recognition device in a wired or wireless manner, and when the wired connection manner is adopted, a light signal line, a power line and the like can extend upwards along the non-use surface of the laryngoscope lens 110, enter the connecting shell 120 and finally enter the handle 500.

The image processing identification means may be built into the handle 500 to increase the portability of the laryngoscope. Two batteries 510 can be arranged in the handle 500 as shown in fig. 4 to supply power for the negative pressure suction motor 210, the camera 400, the image processing and identifying device and other electric devices, thereby improving the portability of the laryngoscope. The handle 500 is provided with a detachable rear cover 520 for removing the built-in battery 510 and the like. The inner end surface of the rear cover 520 may further be provided with a support ring 521 for pressing the battery 510 so that the battery 510 is tightly mounted and fastened without shaking, and the support ring 521 may specifically be a spring. A plurality of second outer teeth 522 can be uniformly arranged on the outer wall surface of the rear cover 520 in the circumferential direction, so that the friction force of screwing is increased, and screwing type disassembly is facilitated.

When the opening of the vacuum suction motor 210 is manual or a manual control switch is required to be provided for the vacuum suction motor 210, the switch can be arranged on the handle 500, the on-off of the vacuum can be controlled by controlling the power supply of the battery 510, even the magnitude of the vacuum can be further controlled by a knob switch, and in addition, the on-off control of the power supply of all the electric equipment of the portable laryngoscope can be realized by the switch.

In one embodiment, the barrel of the liquid cartridge 200 is provided with a through hole, the through hole is provided with an H-shaped sealing ring 220, and the other end of the suction hose 300 penetrates through a center hole of the H-shaped sealing ring 220 and is in clearance fit with the center hole.

The H-shaped sealing ring 220 means that the positions of the outer wall surface of the sealing ring close to the two end surfaces are provided with extending edges along the radial outward direction, and the extending edges at the two ends clamp the hole wall at the through hole of the connecting shell 120 in the connecting shell, so that the sealing ring is stably installed. The imbibition hose 300 is clamped in the central hole of the sealing ring 220, and the diameter of the hole wall of the central hole in a natural state is slightly smaller than the outer diameter of the imbibition hose 300 so as to clamp the imbibition hose 300 and avoid axial shaking of the imbibition hose 300 and separation from the liquid storage cylinder 200.

In one embodiment, the end of the cartridge 200 connected to the handle 500 is provided with a connection cap 230, and the cartridge 200 is threadedly connected to the handle 500 through the connection cap 230.

As shown in fig. 4 and 5, an extending edge extends outward along the axial direction at the outer edge of the connecting end cap 230, an inner wall of the extending edge is provided with an internal thread, and an outer wall of the end portion of the handle 500 is provided with an external thread matching with the internal thread to realize a threaded connection. It is understood that the handle 500 and the liquid storage barrel 200 may be any detachable connection manner suitable for the scene, such as a snap connection, a groove connection, a clamp connection, a snap connection, and the like.

Under the mode that adopts above-mentioned threaded connection, both be airtight space in order to realize liquid storage barrel 200 inner space, can conveniently open liquid storage barrel 200 again in order to clear up, discharge the stock solution to its inside, connect end cover 230 can be set up to and be detachable between the liquid storage barrel 200, connect end cover 230 through detachable connected mode such as spiro union and be connected with liquid storage barrel 200 barrel, the installation in order to realize portable laryngoscope on the spiro union handle 500 again. When the liquid storage cylinder 200 needs to be detached for cleaning, the connecting end cover 230 can be detached, so that the inner space of the connecting end cover 230 is not closed any more.

In one embodiment, the laryngoscope blade assembly 100 is hinged to the reservoir 200.

As shown in fig. 6, the end surface of the liquid storage barrel 200 is provided with two lugs 250 as hinges, the connecting shell 120 of the laryngoscope blade assembly 100 is provided with a through hole for the hinge shaft 260 to pass through, the hinge shaft 260 passes through the connecting shell 120 and is connected with the two lugs 250, so that the connecting shell 120 and the liquid storage barrel 200 can rotate around a rotating shaft which is approximately vertical to the central axis of the handle 500, and the portable laryngoscope is beneficial to saving space when being stored.

In addition, in order to maintain the relative angle between the laryngoscope blade assembly 100 and the handle 500 constant, a spring and hook hole structure (not shown) may be provided between the laryngoscope blade assembly 100 and the handle 500, the spring and hook hole structure being provided on the same side of the hinge shaft 260 in the radial direction and both being located on the side from which the laryngoscope blade extends.

When the laryngoscope is in a folded state, the spring is in a compressed state, the hook of the hook hole structure hooks the hole on the liquid storage cylinder 200, the hook hole structure overcomes the restoring force of the spring (the outward thrust of the spring when the spring returns to the normal state) to form a smaller included angle between the laryngoscope blade 110 and the handle 500, the restoring force at the moment is born by the hook hole structure, and the laryngoscope blade 110 and the handle 500 are folded to realize the space saving under the storage state.

When the laryngoscope needs to be opened, the hook is disengaged from the hole on the liquid storage cylinder 200, the laryngoscope blade assembly 100 loses the constraint of the hook hole structure and rotates around the hinge shaft 260 under the restoring force of the spring, the included angle between the laryngoscope blade 110 and the handle 500 along with the rotation is enlarged to a right angle or an expected angle, at the moment, one side of the connecting shell 120 is abutted against the end surface of the liquid storage cylinder 200, so that the spring cannot further push the laryngoscope blade assembly 100 to rotate, the spring continues to apply pushing force to the connecting shell 120 with smaller restoring force, and the restoring force at the moment is borne by the liquid storage cylinder 200 so as to keep the included angle between the laryngoscope blade 110 and the handle 500.

If the included angle between the laryngoscope blade 110 and the handle 500 can be changed at any time during the use process without effort, the hinge shaft 260 can be set as a screw, a nut is arranged at the tail end of one end of the screw, and the holes on the double lugs 250 and the connecting shell 120 are all unthreaded holes. In a normal state, the screw penetrates through the unthreaded hole, and the connecting shell 120 can rotate freely. When the included angle between the laryngoscope blade 110 and the handle 500 needs to be fixed, the nut is screwed to generate friction between the connecting shell 120 and the side wall surfaces of the two lugs 250, so that the connecting shell 120 is clamped and cannot rotate. This also enables the laryngoscope to be folded and unfolded and the laryngoscope blade to be captured in any angular position. A first embodiment of the disclosed method of detection of residual laryngeal fluid for laryngoscopes is described in detail below with reference to fig. 7. As shown in fig. 7, the method disclosed in this embodiment is implemented based on any one of the first to fourth embodiments of the portable laryngoscope described above. The method comprises step 1.

Step 1, identifying a liquid area from a throat image collected by an image collecting device which is arranged near a liquid suction port at the tail end of a laryngoscope and is positioned at the tail end of the lens, and controlling a liquid suction device to stop sucking liquid according to the area of the liquid area. Wherein the liquid suction device is used for sucking throat liquid into the liquid suction port.

In one embodiment, step 1 comprises step 10.

Step 10, identifying at least one pixel point belonging to a liquid region from the throat image, merging the pixel points in the throat image which have the same image characteristics as the at least one pixel point according to the image characteristics of the pixel points in the liquid region to identify the liquid region in the throat image, and then controlling the liquid suction device to stop liquid suction according to the area of the liquid region.

As shown in fig. 8, in one embodiment, the step 10 includes steps 100 through 500.

And step 100, performing window traversal on the throat image, and calculating the area color mean value of each window area.

And 200, taking a coordinate point in a region corresponding to the region color mean value with the minimum difference value with the color reference value as a seed point, and taking the seed point as an initial liquid region pixel set.

Step 300, calculating a difference value between a color value of each adjacent pixel point of the region formed by the liquid region pixel set and a color mean value of the liquid region pixel set, and adding the adjacent pixel point corresponding to the difference value smaller than a color difference threshold value into the liquid region pixel set so as to update the liquid region pixel set.

Step 400, performing iteration of the liquid pixel judgment step on the updated liquid region pixel set until the difference values corresponding to all adjacent pixel points of the liquid region pixel set are not less than the color difference threshold value, and then taking the latest liquid region pixel set as a liquid region.

And 500, controlling a liquid suction device to stop sucking liquid when the area of the liquid area is lower than a preset area threshold value.

As shown in fig. 9, in one embodiment, the color reference value includes a first reference value corresponding to a first color liquid and a second reference value corresponding to a second color liquid. And obtaining a first liquid area pixel set and a second liquid area pixel set based on the first reference value and the second reference value respectively, updating the first liquid area pixel set and the second liquid area pixel set respectively to obtain a first liquid area corresponding to the first color liquid and a second liquid area corresponding to the second color liquid respectively, and controlling the liquid suction device to stop sucking liquid when the area of the first liquid area is lower than the area threshold of the first color liquid and the area of the second liquid area is lower than the area threshold of the second color liquid.

In one embodiment, the method further comprises: firstly, graying the collected throat image to obtain a gray image. And the throat image subjected to window traversal is the gray image, and the color value, the color mean value, the color reference value and the color difference threshold value adopt corresponding gray values.

In an embodiment, if there are a plurality of regions corresponding to the color mean value of the region with the minimum difference from the color reference value, one of the regions is selected as a region for selecting a seed point, or the color mean value of each adjacent region of each of the regions is calculated, and the region corresponding to the color mean value of the adjacent region with the minimum difference from the color reference value is used as the region for selecting a seed point.

In one embodiment, when updating the set of liquid region pixels, a color mean of the set of liquid region pixels is also calculated and updated.

In one embodiment, the pipetting device is further controlled to initiate pipetting when the liquid region area is not below the area threshold, and the pipetting power is adjusted in dependence on the difference between the liquid region area and the area threshold.

In the description of the present application, it is to be understood that the terms "central," "longitudinal," "lateral," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," and the like are used in the orientation or positional relationship indicated in the drawings, which are intended to be based on the orientation or positional relationship shown in the drawings, and are used merely for convenience in describing the present application and to simplify the description, but do not indicate or imply that the referenced device or element must have a particular orientation, be constructed and operated in a particular orientation, and therefore should not be taken as limiting the scope of the present application.

The division of the modules herein is merely a division of logical functions, and other divisions may be possible in actual implementation, for example, a plurality of modules may be combined or integrated in another system. Modules described as separate components may or may not be physically separate. Therefore, some or all of the units can be selected according to actual needs to implement the scheme of the embodiment.

The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present application should be covered within the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.