阅读说明：本技术 B型主动脉夹层术后风险预测方法、装置和电子设备 (Method and device for predicting risk after B-type aortic dissection operation and electronic equipment ) 是由 柴象飞 代双凤 郭伟 郭娜 左盼莉 葛阳阳 于 2019-10-29 设计创作，主要内容包括：本申请实施例提供了一种B型主动脉夹层术后风险预测方法、装置和电子设备,涉及医学影像技术领域。该方法首先获取多例术前三维血管造影图像,对各例所述三维血管造影图像进行预处理以获得待处理数据集,其中,所述待处理数据集包括血管真腔信息及血管假腔信息。接着,对所述血管真腔信息及所述血管假腔信息进行矢量分析,获得多种影像组学特征信息,并对多种影像组学特征信息进行降维处理及线性拟合,得到风险因素变量。最后,根据风险因素变量及预设临床病变类型,构建风险预测模型。如此,可获得术后发生并发症的风险概率,从而保障患者安全。(The embodiment of the application provides a method and a device for predicting risks after a B-type aortic dissection operation and electronic equipment, and relates to the technical field of medical images. The method comprises the steps of firstly obtaining a plurality of preoperative three-dimensional angiography images, and preprocessing each three-dimensional angiography image to obtain a data set to be processed, wherein the data set to be processed comprises blood vessel true lumen information and blood vessel false lumen information. And then, carrying out vector analysis on the blood vessel true lumen information and the blood vessel false lumen information to obtain various image omics characteristic information, and carrying out dimension reduction processing and linear fitting on the various image omics characteristic information to obtain a risk factor variable. And finally, constructing a risk prediction model according to the risk factor variable and the preset clinical lesion type. Therefore, the risk probability of postoperative complications can be obtained, and the safety of patients is guaranteed.)

1. A method for predicting risk after type B aortic dissection, the method comprising:

acquiring a plurality of preoperative three-dimensional angiography images, and preprocessing each three-dimensional angiography image to obtain a data set to be processed, wherein the data set to be processed comprises blood vessel true lumen information and blood vessel false lumen information;

performing vector analysis on the blood vessel true lumen information and the blood vessel false lumen information to obtain various image omics characteristic information, and performing dimension reduction processing and linear fitting on the various image omics characteristic information to obtain a risk factor variable;

and constructing a risk prediction model according to the risk factor variable and a preset clinical lesion type.

2. The method for predicting risk after aortic dissection according to claim 1, wherein the step of performing vector analysis on the blood vessel true lumen information and the blood vessel false lumen information to obtain a plurality of image group characteristic information comprises:

calculating a matching point pair between a false lumen image and a real lumen image contained in the three-dimensional angiography image according to the blood vessel real lumen information and the blood vessel false lumen information;

obtaining a vector of the matching point pair, and calculating a direction angle of the vector;

and performing image omics feature extraction on the vector and the direction angle to obtain various image omics feature information.

3. The method according to claim 2, wherein the step of calculating the matching point pairs between the false lumen image and the true lumen image included in the three-dimensional angiography image according to the blood vessel true lumen information and the blood vessel false lumen information comprises:

acquiring a first range of a real cavity image contained in the three-dimensional angiography image in a preset direction according to the blood vessel real cavity information, and sampling on a geometric central line of the real cavity image according to the first range and a first preset interval to acquire a plurality of first characteristic points;

according to the blood vessel false lumen information, obtaining a second range of a false lumen image contained in the three-dimensional angiography image in the preset direction, and according to the second range, sampling on a geometric central line of the false lumen image according to a second preset interval to obtain a plurality of second feature points;

and performing characteristic point matching on the first characteristic point and the second characteristic point to obtain a plurality of matched point pairs.

4. The method for predicting risk after aortic dissection according to claim 1, wherein the step of performing dimension reduction processing and linear fitting on the multiple kinds of image omics characteristic information to obtain the risk factor variable comprises:

standardizing the characteristic information of the multiple image omics to obtain standardized characteristic information;

performing dimension reduction processing on the standardized feature information based on a preset regression model to obtain a risk factor feature after the dimension reduction processing and a correlation coefficient corresponding to the risk factor feature;

and performing linear fitting on each risk factor characteristic according to the corresponding correlation coefficient to obtain a risk factor variable.

5. The method for predicting risk after aortic dissection according to claim 4, wherein the step of linearly fitting each risk factor feature according to its corresponding correlation coefficient to obtain a risk factor variable comprises:

and for each risk factor characteristic, performing linear fitting on the risk factor characteristic according to a corresponding correlation coefficient thereof according to the following formula to obtain a risk factor variable:

wherein Radscore is the risk factor variable, RadF_iCoef being characteristic of said risk factor_iAnd the correlation coefficient corresponding to the risk factor characteristic.

6. The method for predicting risk after aortic dissection according to claim 1, wherein the step of preprocessing the three-dimensional angiography images to obtain the data set to be processed comprises:

and labeling each three-dimensional angiography image according to a preset labeling mode according to the blood vessel structure of the human body to obtain a data set to be processed.

7. The method for predicting risk after aortic dissection of type B according to claim 1, wherein after constructing the risk prediction model, the method further comprises:

obtaining an original three-dimensional angiography image;

obtaining a lesion type and a risk factor value corresponding to a risk factor variable according to the original three-dimensional angiography image;

and calculating the postoperative risk probability according to the risk factor value, the lesion type and the risk prediction model to obtain a risk prediction result.

8. A post-aortic dissection risk prediction device of type B, comprising:

the preprocessing module is used for acquiring a plurality of preoperative three-dimensional angiography images and preprocessing each three-dimensional angiography image to obtain a data set to be processed, wherein the data set to be processed comprises blood vessel true lumen information and blood vessel false lumen information;

the characteristic analysis module is used for carrying out vector analysis on the blood vessel true lumen information and the blood vessel false lumen information to obtain a plurality of types of image omics characteristic information, and carrying out dimension reduction processing and linear fitting on the plurality of types of image omics characteristic information to obtain a risk factor variable;

and the construction module is used for constructing a risk prediction model according to the risk factor variable and the clinical lesion type.

9. An electronic device, comprising a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate via the bus, and the processor executes the machine-readable instructions to perform the steps of the method for predicting risk after aortic dissection according to any one of claims 1 to 7.

10. A readable storage medium, wherein a computer program is stored in the readable storage medium, which computer program, when executed, implements the post-aortic dissection risk prediction method of any one of claims 1-7.

Technical Field

The application relates to the field of medical imaging, in particular to a method and a device for predicting risk after B-type aortic dissection and electronic equipment.

Background

Type B Aortic Dissection (TBAD) is a true or false lumen formed by tearing of the Aortic intima, which allows blood flow to enter the media of the Aortic wall through a tear to separate the media and expand along the major axis of the aorta.

Currently, Thoracic Aortic endoluminal Repair (TEVAR) is mainly used for treatment of type B Aortic dissections. The basic principle is that the covered stent covers the first laceration, the blood supply of the false cavity is isolated, the blood supply of the true cavity is recovered, complete thrombosis in the false cavity is expected, and finally good artery remodeling is finished. However, TEVAR treatment is not once and for all and is often followed by complications such as thoracic aortic dilation, abdominal aortic dilation, etc. after surgery. Patients with higher severity of complications still require secondary surgical intervention, which can significantly increase patient pain and social medical burden.

Therefore, how to know the risk probability of occurrence of the TEVAR postoperative complications so as to ensure the safety of patients is a problem which needs to be solved urgently at present.

Disclosure of Invention

In view of the above, embodiments of the present application provide a method, an apparatus, and an electronic device for predicting risk after B-aortic dissection to solve the above problems.

The embodiment of the application can be realized as follows:

in a first aspect, an embodiment of the present application provides a method for predicting risk after a type B aortic dissection operation, the method including:

acquiring three-dimensional angiography data before a plurality of cases of operations, and preprocessing the three-dimensional angiography data to obtain a data set to be processed, wherein the data set to be processed comprises blood vessel true cavity information and blood vessel false cavity information;

performing vector analysis on the blood vessel true lumen information and the blood vessel false lumen information to obtain various image omics characteristic information, and performing dimension reduction processing and linear fitting on the various image omics characteristic information to obtain a risk factor variable;

and constructing a risk prediction model according to the risk factor variable and a preset clinical lesion type.

In an optional embodiment, the step of performing vector analysis on the blood vessel true lumen information and the blood vessel false lumen information to obtain a plurality of types of imaging group feature information includes:

obtaining a matching point pair between a false lumen image and a real lumen image contained in the three-dimensional angiography image according to the blood vessel real lumen information and the blood vessel false lumen information;

obtaining a vector of the matching point pair, and calculating a direction angle of the vector;

and performing image omics feature extraction on the vector and the direction angle to obtain various image omics feature information.

In an optional embodiment, the step of calculating a matching point pair between a false lumen image and a true lumen image included in the three-dimensional angiography image according to the blood vessel true lumen information and the blood vessel false lumen information includes:

acquiring a first range of a real cavity image contained in the three-dimensional angiography image in a preset direction according to the blood vessel real cavity information, and sampling on a geometric central line of the real cavity image according to the first range and a first preset interval to acquire a plurality of first characteristic points;

according to the blood vessel false lumen information, obtaining a second range of a false lumen image contained in the three-dimensional angiography image in the preset direction, and according to the second range, sampling on a geometric central line of the false lumen image according to a second preset interval to obtain a plurality of second feature points;

and performing characteristic point matching on the first characteristic point and the second characteristic point to obtain a plurality of matched point pairs.

In an optional embodiment, the step of performing dimension reduction processing and linear fitting on the multiple kinds of imaging group feature information to obtain the risk factor variable includes:

standardizing the characteristic information of the multiple image omics to obtain standardized characteristic information;

performing dimension reduction processing on the standardized feature information based on a preset regression model to obtain a risk factor feature after the dimension reduction processing and a correlation coefficient corresponding to the risk factor feature;

and performing linear fitting on each risk factor characteristic according to the corresponding correlation coefficient to obtain a risk factor variable.

In an optional embodiment, the step of performing linear fitting on each risk factor characteristic according to the corresponding correlation coefficient to obtain a risk factor variable includes:

and for each risk factor characteristic, performing linear fitting on the risk factor characteristic according to a corresponding correlation coefficient thereof according to the following formula to obtain a risk factor variable:

wherein Radscore is the risk factor variable, RadF_iCoef being characteristic of said risk factor_iAnd the correlation coefficient corresponding to the risk factor characteristic.

In an alternative embodiment, the step of preprocessing the three-dimensional angiography data to obtain a to-be-processed data set includes:

labeling the three-dimensional angiography data according to a preset labeling mode according to the structure of the blood vessel of the human body, manually segmenting the blood vessel true-false cavity, and obtaining a data set to be processed.

In an alternative embodiment, after the risk prediction model is constructed, the method further comprises:

obtaining original three-dimensional angiography data;

obtaining a lesion type and a risk factor value corresponding to a risk factor variable according to the three-dimensional original angiography data;

and calculating the postoperative risk probability according to the risk factor value, the lesion type and the risk prediction model to obtain a risk prediction result.

In a second aspect, an embodiment of the present application provides a post-aortic dissection risk prediction apparatus for type B aorta, including:

the preprocessing module is used for acquiring a plurality of preoperative three-dimensional angiography images and preprocessing each three-dimensional angiography image to obtain a data set to be processed, wherein the data set to be processed comprises blood vessel true lumen information and blood vessel false lumen information;

the characteristic analysis module is used for carrying out vector analysis on the blood vessel true lumen information and the blood vessel false lumen information to obtain a plurality of types of image omics characteristic information, and carrying out dimension reduction processing and linear fitting on the plurality of types of image omics characteristic information to obtain a risk factor variable;

and the construction module is used for constructing a risk prediction model according to the risk factor variable and the clinical lesion type.

In a third aspect, an embodiment of the present application provides an electronic device, including a processor, a memory and a bus, where the memory stores machine-readable instructions executable by the processor, and when the electronic device runs, the processor and the memory communicate with each other through the bus, and the processor executes the machine-readable instructions to perform the steps of the method for predicting risk after aortic dissection.

In a fourth aspect, the present application provides a readable storage medium, in which a computer program is stored, and the computer program is executed to implement the method for predicting risk after B-type aortic dissection according to any one of the foregoing embodiments.

The embodiment of the application provides a method and a device for predicting risks after a type-B aortic dissection operation and an electronic device. The method comprises the steps of firstly obtaining three-dimensional angiography data before a plurality of cases of operations, and preprocessing the three-dimensional angiography data to obtain a data set to be processed, wherein the data set to be processed comprises blood vessel true lumen information and blood vessel false lumen information. And then, carrying out vector analysis on the blood vessel true lumen information and the blood vessel false lumen information to obtain various image omics characteristic information, and carrying out dimension reduction processing and linear fitting on the various image omics characteristic information to obtain an image risk factor variable. And finally, constructing a risk prediction model according to the risk factor variable and a preset clinical lesion type. Therefore, the risk probability of postoperative complications can be obtained, and the safety of patients is guaranteed.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained from the drawings without inventive effort.

Fig. 1 is a block diagram of an electronic device according to an embodiment of the present disclosure;

fig. 2 is a flowchart of a risk prediction method after a type B aortic dissection provided in an embodiment of the present application;

fig. 3 is a flowchart illustrating one of sub-steps of a method for predicting risk after a type B aortic dissection according to an embodiment of the present disclosure;

FIG. 4 is a schematic view of a vector projection provided by an embodiment of the present application;

fig. 5 is a second flowchart illustrating sub-steps of a method for predicting risk after aortic dissection in type B according to an embodiment of the present disclosure;

FIG. 6 is a risk prediction model provided in an embodiment of the present application;

FIG. 7 is a graph illustrating a correction curve of a risk prediction model provided in an embodiment of the present application;

FIG. 8 is a decision graph of a risk prediction model corresponding to three cases provided by an embodiment of the present application;

fig. 9 is a functional block diagram of a risk prediction apparatus according to an embodiment of the present application.

Icon: 100-an electronic device; 110-a memory; 120-a processor; 130-risk prediction means; 131-a pre-processing module; 132-a feature analysis module; 133-building a module.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments of the present application, but not all embodiments. The components of the embodiments of the present application, generally described and illustrated in the figures herein, can be arranged and designed in a wide variety of different configurations.

Thus, the following detailed description of the embodiments of the present application, presented in the accompanying drawings, is not intended to limit the scope of the claimed application, but is merely representative of selected embodiments of the application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.

Furthermore, the appearances of the terms "first," "second," and the like, if any, are used solely to distinguish one from another and are not to be construed as indicating or implying relative importance.

It should be noted that the features of the embodiments of the present application may be combined with each other without conflict.

Aortic dissection type B refers to dissection without involvement of the ascending aorta, with high mortality rate, and about 10% of patients die during hospitalization, and recent system reviews report that 10% -54% of patients suffer thoracic aortic dilation after TEVAR surgery, and persistent arterial dilation leads to dissection aneurysm, increasing the risk of advanced secondary surgery and death among patients.

The inventor finds that the risk factor of thoracic aorta dilatation after TEVAR operation is a branch artery supplying blood to a thoracic cavity, the number of the branch arteries supplying blood to the thoracic cavity is closely related to the spatial position relationship of a thoracic cavity true and false cavity, and the spatial position relationship of the thoracic cavity true and false cavity is a substitute index of the branch artery supplying blood to the thoracic cavity. According to the spatial position relation of the true and false cavities of the thoracic section, currently, the clinical medicine proposes that B-type aortic dissection disease is classified as 301 classification, and the 301 classification is also the main basis for thoracoaortic endoluminal repair. However, such a parting still has certain drawbacks: the classification is not beneficial to accurately judging the thoracic aorta dilation risk of the interbedded individual, the classification method is complex in judgment standard, interpretation needs to be carried out layer by layer according to the axial position of a Computed Tomography (CTA) image, and the consistency among interpreters is not ideal.

Therefore, a method for accurately predicting the risk after the TEVAR operation of the type B aortic dissection is needed, and the risk probability of the occurrence of complications after the TEVAR operation is known, so that the safety of a patient is ensured.

Based on the above findings, the embodiment of the present application provides a method, an apparatus, and an electronic device for predicting a risk after B-aortic dissection. The method is characterized in that a risk prediction model is jointly constructed by combining the relative position relation of true and false cavities of the aortic dissection and clinically known lesion types (301 typing) so as to evaluate and predict the risk probability of the occurrence of the TEVAR postoperative complications. The above process is described in detail with reference to specific examples.

Referring to fig. 1, fig. 1 is a schematic structural diagram of an electronic device 100 according to an embodiment of the present disclosure. The apparatus may include a processor 120, a memory 110, a risk prediction device 130, and a bus, wherein the memory 110 stores machine-readable instructions executable by the processor 120, when the electronic apparatus 100 is operated, the processor 120 and the memory 110 communicate with each other through the bus, and the processor 120 executes the machine-readable instructions and performs the steps of the post-aortic dissection risk prediction method.

The memory 110, the processor 120, and other components are electrically connected to each other directly or indirectly to enable signal transmission or interaction.

For example, the components may be electrically connected to each other via one or more communication buses or signal lines. The risk prediction means 130 comprises at least one software function which may be stored in the memory 110 in the form of software or firmware. The processor 120 is configured to execute executable modules stored in the memory 110, such as software functional modules or computer programs included in the risk prediction unit 130.

The Memory 110 may be, but is not limited to, a Random Access Memory (RAM), a Read Only Memory (ROM), a Programmable Read-Only Memory (PROM), an Erasable Read-Only Memory (EPROM), an electrically Erasable Read-Only Memory (EEPROM), and the like.

The processor 120 may be an integrated circuit chip having signal processing capabilities. The processor 120 may be a general-purpose processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and so on. But may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components. The various methods, steps, and logic blocks disclosed in the embodiments of the present application may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

In the embodiment of the present application, the memory 110 is used for storing a program, and the processor 120 is used for executing the program after receiving the execution instruction. The method defined by the process disclosed in any of the embodiments of the present application can be applied to the processor 120, or implemented by the processor 120.

In the embodiment of the present application, the electronic device 100 may be, but is not limited to, a Personal Computer (PC), a tablet computer, or other device with a processing function.

It will be appreciated that the configuration shown in figure 1 is merely illustrative. Electronic device 100 may also have more or fewer components than shown in FIG. 1, or a different configuration than shown in FIG. 1. The components shown in fig. 1 may be implemented in hardware, software, or a combination thereof.

Referring to fig. 2, fig. 2 is a flowchart illustrating a risk prediction method after a type B aortic dissection according to an embodiment of the present disclosure. The specific flow shown in fig. 2 is described in detail below.

S1, acquiring a plurality of preoperative three-dimensional angiography data, and preprocessing the three-dimensional angiography images to obtain a data set to be processed, wherein the data set to be processed comprises blood vessel true lumen information and blood vessel false lumen information.

In the embodiment of the application, the selected three-dimensional angiography data is aorta CTADICOM data before B-type aortic dissection. These images were all from a two-way, multi-center clinical trial from the general hospital of the liberation force, which included TBAD patients receiving Thoracic aortic endoluminal Repair (TEVAR) treatment from 2011, 1 month, to 2012, 12 months, 31 days.

CTA is also called non-invasive blood vessel imaging technology, is a reconstruction method performed after enhanced CT scanning, which is simply to inject a certain amount of contrast agent into a blood vessel during scanning to fill the blood vessel cavity and acquire data of the blood vessel cavity, and then to perform computer reconstruction processing to delete the content of a non-interest region, only retain the image of the blood vessel and perform overall and multi-angle reconstruction, so that the overall appearance of the blood vessel is fully displayed, and simultaneously, the display of a diseased blood vessel is facilitated.

Optionally, in this embodiment of the present application, the step of preprocessing the three-dimensional angiography image to obtain the to-be-processed data set may be:

first, data cleaning is performed on the three-dimensional angiographic image.

Then, as an implementation manner, the three-dimensional angiography images can be labeled according to the blood vessel structure of the human body in a preset labeling manner to obtain a data set to be processed.

As an embodiment, the type B aortic dissection may be segmented according to the vascular structure of the human body, and the labeling may range from the initial upper edge of the left coronary artery to the bifurcation of the abdominal aorta, including the ascending aorta, the aortic arch and the descending aorta. The whole aorta was labeled with label Lable1 (green), the true lumen with label Lable2 (yellow) and the false lumen with label Lable3 (red).

The color corresponding to each label in the preset labeling mode can be other colors, and only the colors need to be distinguished.

As another implementation mode, the blood vessel real-false cavity can be segmented according to the blood vessel structure of the human body, and the three-dimensional angiography images are labeled according to a preset labeling mode.

And S2, performing vector analysis on the blood vessel true lumen information and the blood vessel false lumen information to obtain a plurality of kinds of image omics characteristic information, and performing dimension reduction processing and linear fitting on the plurality of kinds of image omics characteristic information to obtain a risk factor variable.

Alternatively, referring to fig. 3 and 4 in combination, the blood vessel true lumen information and the blood vessel false lumen information may be subjected to vector analysis through the steps S21-S23 shown in fig. 3 to obtain a plurality of types of imaging characteristic information.

Fig. 4 is a schematic view of the vector projection obtained after steps S21-S23, in which the thicker black line is the geometric centerline of the true lumen of the blood vessel, and the thinner line is the geometric centerline of the false lumen of the blood vessel.

And S21, obtaining matching point pairs between the false lumen image and the true lumen image contained in the three-dimensional angiography image according to the blood vessel true lumen information and the blood vessel false lumen information.

Firstly, according to the blood vessel true lumen information, a first range of a true lumen image contained in the three-dimensional angiography image in a preset direction is obtained, according to the first range, sampling is carried out on a geometric central line of the true lumen image according to a first preset interval, and a plurality of first feature points are obtained.

And then, according to the blood vessel false lumen information, obtaining a second range of a false lumen image contained in the three-dimensional angiography image in the preset direction, and according to the second range, sampling on a geometric central line of the false lumen image according to a second preset interval to obtain a plurality of second feature points.

As an embodiment, optionally, first, the plurality of first feature points may be acquired by obtaining a longitudinal extent of the true lumen of the blood vessel as the first extent, and sampling equidistantly along a geometric centerline of the true lumen of the blood vessel. Then, a second range corresponding to the blood vessel false cavity in the longitudinal direction can be obtained according to the longitudinal range of the blood vessel false cavity. And sampling equidistantly along the geometric central line of the blood vessel false cavity to obtain a plurality of second characteristic points.

And finally, performing characteristic point matching on the first characteristic point and the second characteristic point to obtain a plurality of matched point pairs.

And S22, obtaining the vector of the matching point pair, and calculating the direction angle of the vector.

Obtaining the vector corresponding to each matching point pair, and calculating the direction angle of the vector, specifically, the calculation mode of the direction angle may refer to a corresponding solution method in the spatial resolution geometry, which is not described herein again.

And S23, performing image omics feature extraction on the vector and the direction angle to obtain various image omics feature information.

Alternatively, the features may be extracted using pyradiomics to obtain a plurality of first-order grayscale features, such as entropy, kurtosis, absolute deviation, variance, skewness, r-order moment, r-order center moment, and torsion degree, etc., about a plurality of vectors and direction angles.

Referring to fig. 5, the multiple image omics characteristic information is subjected to dimension reduction and linear fitting through steps S24-S26 to obtain risk factor variables.

And S24, standardizing the characteristic information of the multiple image omics to obtain standardized characteristic information.

Alternatively, Z-score (ZScore) can be used to normalize the various imaging omics signature information. That is, the standard deviation a of each feature information is calculated. Next, a difference B between each piece of feature information and the average value of each piece of feature information is obtained. And finally, obtaining a quotient of the difference B and the standard deviation A, namely the standardized feature information obtained after the feature information is subjected to standardized processing.

And S25, performing dimension reduction processing on the standardized feature information based on a preset regression model, and obtaining the risk factor features after the dimension reduction processing and the correlation coefficients corresponding to the risk factor features.

Optionally, in the embodiment of the present application, the LASSO-Cox algorithm performs a proportional risk model analysis on the normalized feature information.

And testing the accuracy of the to-be-selected risk factor characteristics by using cross-folding verification (10-fold cross-validation), and selecting the to-be-selected risk factor corresponding to the value with the minimum error value as the risk factor characteristics after the dimension reduction processing, so as to obtain a plurality of risk factor characteristics after the dimension reduction processing and the correlation coefficient corresponding to the risk factor characteristics.

As shown in table 1. Wherein, X10Percent is 10 percentile, Encopy is Entropy, Kurtosis is Kurtosis, Max is maximum, robustmeans Absolute Deviation is robust average value (average value of gray scale between 10 percentile and 90 percentile), RootMeanSquare is square of root value, Skewness is Skewness, and Variance is Variance.

And S26, performing linear fitting on each risk factor characteristic according to the corresponding correlation coefficient to obtain a risk factor variable.

And for each risk factor characteristic, performing linear fitting on the risk factor characteristic according to a corresponding correlation coefficient thereof according to the following formula to obtain a risk factor variable:

wherein Rad _ score is the risk factor variable, RadF_iCoef being characteristic of said risk factor_iAnd the correlation coefficient corresponding to the risk factor characteristic.

TABLE 1 Risk factor characteristics after dimension reduction and correlation coefficients corresponding to the risk factor characteristics

And S3, constructing a risk prediction model according to the risk factor variables and the preset clinical lesion types.

Optionally, as shown in fig. 6, using the risk factor variables and preset clinical lesion types, a Nomogram (Nomogram) is constructed as a risk prediction model, so that the prediction model is visualized.

Wherein Points is a variable integral, Total Points is a Total score, and the calconfiguration _301 is a preset lesion type (i.e., type-B aortic dissection lesion type 301 suggested by the 301 hospital). Rad _ score is a risk factor variable. 1-year thoracic aortic expansion probability, 2-year thoracic aortic expansion probability, and 3-year thoracic aortic expansion probability, respectively.

Meanwhile, as an implementation method, as shown in fig. 7, the calibration degree of the risk prediction model may be evaluated using a calibration curve. Wherein, the dotted line is a standard value, the dot-dash line is a probability value of aortic dilatation occurring within 3 years, the short line is a probability value of aortic dilatation occurring within 2 years, and the solid line is a probability value of aortic dilatation occurring within 1 year. As can be seen from fig. 7, the probability values of aortic dilatation occurring in each year in the nomogram have better agreement than the norm.

As another implementation method, a decision curve can be used to verify the effectiveness of the risk prediction model. Referring to fig. 8, fig. 8 shows a Logistic regression model respectively performed on the lesion type (301 type), risk factor variable, and combined lesion type (301 type) and risk factor variable of the aortic dissection to obtain a risk prediction model through fitting, so as to draw decision curves under three conditions.

In the figure, a curve formed by a dotted line is a multi-factor model decision curve corresponding to Rad _ score and Classification (combined lesion type (301 type) and risk factor variable), a curve formed by a dotted line is a risk prediction model decision curve corresponding to a single factor of Rad _ score (risk factor variable), and a solid line is a decision curve of a risk prediction model corresponding to a single factor of which the curve is based on Classification (lesion type). As can be seen from the figure, the risk prediction model based on the risk factor variable is more advantageous in predicting the postoperative complications than the risk prediction model based on the lesion type, and the multi-factor model based on the combination of the lesion type (301 classification) and the risk factor variable is more advantageous in predicting the occurrence probability of the postoperative complications.

Optionally, after the risk prediction model is constructed, the risk prediction result can be obtained by the following method when the risk prediction model is actually used clinically:

first, three-dimensional raw angiographic data is obtained.

And then, obtaining the clinical lesion type and a risk factor value corresponding to the risk factor variable according to the three-dimensional original angiography image.

And finally, calculating the postoperative risk probability according to the risk factor value, the lesion type and the risk prediction model to obtain a risk prediction result.

Referring again to fig. 6, the risk prediction model (nomogram) is used as follows: the risk factor value and the lesion type respectively correspond to corresponding variable Points (indicating the uppermost Points), then each variable point is added to obtain a Total point, and the score of the Total point corresponds to the probability value of thoracic aortic expansion occurring within the lowest 1 year, the probability value of thoracic aortic expansion occurring within 2 years or the probability value of thoracic aortic expansion occurring within 3 years, so that the risk prediction results corresponding to different years can be known.

The embodiment of the application provides a risk prediction method after a B-type aortic dissection operation, which is characterized in that a risk prediction model is established by extracting characteristic information in an angiography image of an aorta and combining the characteristic information with a clinically known lesion type, so that the risk probability of complications after a TEVAR operation can be known, the TBAD prognosis prediction level can be improved, a TEVAR treatment strategy can be optimized, the long-term prognosis of a patient can be improved, and a method reference can be provided for the exploration application of an imaging omics technology in the field of aortic diseases.

Referring to fig. 9, the present embodiment also provides a risk prediction apparatus 130, including:

the preprocessing module 131 is configured to acquire a plurality of preoperative three-dimensional angiography images, and preprocess the three-dimensional angiography images to obtain a to-be-processed data set, where the to-be-processed data set includes blood vessel true lumen information and blood vessel false lumen information.

The feature analysis module 132 is configured to perform vector analysis on the blood vessel true lumen information and the blood vessel false lumen information to obtain multiple types of image omics feature information, and perform dimension reduction processing and linear fitting on the multiple types of image omics feature information to obtain a risk factor variable.

And the construction module 133 is configured to construct a risk prediction model according to the risk factor variable and a preset clinical lesion type.

It can be understood that, for the specific operation method of each functional module in the embodiment of the present application, reference may be made to the detailed description of the corresponding step in the foregoing method embodiment, and repeated descriptions are not repeated here.

The embodiment of the application also provides a readable storage medium, wherein a computer program is stored in the readable storage medium, and when the computer program is executed, the computer program realizes the risk prediction method after the B-type aortic dissection.

In summary, the present application provides a method, an apparatus, and an electronic device 100 for predicting risk after B-aortic dissection. The method comprises the steps of firstly obtaining a plurality of preoperative three-dimensional angiography images, and preprocessing each three-dimensional angiography image to obtain a data set to be processed, wherein the data set to be processed comprises blood vessel true lumen information and blood vessel false lumen information. And then, carrying out vector analysis on the blood vessel true lumen information and the blood vessel false lumen information to obtain various image omics characteristic information, carrying out dimensionality reduction processing and linear fitting on the various image omics characteristic information to obtain a risk factor variable, and finally, constructing a risk prediction model according to the risk factor variable and a preset clinical lesion type. Therefore, the risk probability of postoperative complications can be obtained according to the constructed risk prediction model, and the safety of patients is guaranteed.

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.