Poor left ventricular mass regression risk prediction method after transcatheter aortic valve replacement

Abstract

The application discloses a kind of left ventricular mass regression risk prediction method after transcatheter aortic valve replacement, belong to postoperative prediction technical field, its purpose is to solve the technical problem that there is no technology in prior art to effectively predict the risk of left ventricular reverse remodeling after mass regression after transcatheter aortic valve replacement.It includes the following steps: obtaining preoperative heart CT image to be predicted;According to the preoperative heart CT image to be predicted, the image-based parameters of epicardial fat are extracted;According to the extracted image-based parameters, calculate the image-based score;According to the image-based score, calculate the risk probability of left ventricular mass regression after transcatheter aortic valve replacement.Using innovative algorithm, the risk value of left ventricular mass regression can be calculated and predicted, and the risk of left ventricular reverse remodeling after mass regression after transcatheter aortic valve replacement can be effectively predicted.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of postoperative prediction, and relates to aortic valve replacement postoperative risk prediction, in particular to a left ventricular mass regression poor risk prediction method after transcatheter aortic valve replacement. BACKGROUND

[0002] The aortic valve is a structure between the left ventricle and the aorta, which inhibits the backflow of blood flowing into the aorta into the left ventricle. Aortic valve stenosis (AS) is one of the most common heart valve diseases, with a prevalence of 4%-7% in people over 65 years old, and the two-year mortality rate of patients with symptoms can be as high as 50%.

[0003] At present, the effective treatment methods for severe AS include surgical valve replacement and transcatheter aortic valve replacement (TAVR). TAVR is a new minimally invasive treatment technology for aortic valve disease. The surgery releases a compressed artificial aortic valve (a stent with a valve) through a vascular route to the diseased aortic valve to achieve treatment of the disease, and has the characteristics of small trauma, rapid recovery, short hospitalization period, etc.

[0004] In the prior art, research is mostly only conducted on postoperative complication prediction of transcatheter aortic valve replacement, and related complications include paravalvular leakage (blood refluxes back to the left ventricle from the ascending aorta through the gap between the stent and the aortic valve complex) due to the existence of paravalvular gap after the release of the artificial valve, damage to the conduction bundle during the release of the artificial valve, and even permanent pacemaker implantation (atrioventricular block due to compression and damage of the conduction bundle by the artificial valve stent).

[0005] The invention patent with the application number 202011030008.4 discloses a transcatheter aortic valve replacement postoperative complication prediction method, device and equipment. The method comprises the following steps: collecting feature data of a target patient, and inputting the feature data into a preset TAVI postoperative complication prediction model to obtain a TAVI postoperative complication prediction result of the target patient. The TAVI postoperative complication prediction result includes a TAVI postoperative pacemaker implantation prediction result and a TAVI postoperative paravalvular leakage prediction result. In the method, the feature data is obtained from the anatomical index under the CT data, which is easy to obtain and more popular and software-based compared with the finite element scheme. The scheme pays more attention to the initial implantation depth of the artificial aortic valve during the release process, and is more clinically valuable and more accurate in predicting the clinical outcome compared with the finite element scheme, which only focuses on the implantation depth of the artificial valve after the complete release of the valve.

[0006] However, in recent years, more and more studies have shown that increased left ventricular mass is often accompanied by higher all-cause mortality and disease risk. Left ventricular remodeling is a compensatory mechanism of the heart to diseases such as aortic valve stenosis that cause an increase in afterload, and its main manifestation is an increase in left ventricular mass. Studies have confirmed that TAVR can alleviate the increase in left ventricular mass in AS patients, and the left ventricular mass index rapidly decreases by about 14.5%-22% in the first year after TAVR. In addition, a greater degree of left ventricular mass regression at 1 year is associated with lower mortality and hospitalization rates at 5 years after TAVR. Therefore, further elucidation of the predictors of poor postoperative left ventricular mass regression is of great significance for early identification of high-risk patients and improvement of patient outcomes. Epicardial adipose tissue (EAT) is an inflammatory visceral fat depot located within the pericardium, and its imaging and proteomic features on computed tomography (CT) images have been found to be associated with poor cardiac remodeling in AS patients after aortic valve replacement. Considering that preoperative cardiac CT imaging has become a routine examination in contemporary TAVR treatment, it is extremely valuable to use preoperative CT images as a non-invasive method without additional costs to determine the characteristics of EAT. In this context, in addition to predicting the risk of postoperative complications, it is also necessary to predict the risk of poor postoperative left ventricular mass regression after transcatheter aortic valve replacement, especially based on preoperative cardiac CT images. However, there is no technology in the prior art that can effectively predict the risk of poor postoperative left ventricular mass regression after transcatheter aortic valve replacement. SUMMARY

[0007] The purpose of the present application is to solve the technical problem that there is no technology in the prior art that can effectively predict the risk of poor postoperative left ventricular mass regression after transcatheter aortic valve replacement, and to provide a non-invasive, cost-effective, and widely available method for predicting the risk of poor postoperative left ventricular mass regression after transcatheter aortic valve replacement, which can accurately predict the risk of poor postoperative left ventricular mass regression after transcatheter aortic valve replacement.

[0008] The present application specifically adopts the following technical solutions to achieve the above-mentioned purpose:

[0009] A method for predicting the risk of poor postoperative left ventricular mass regression after transcatheter aortic valve replacement, comprising the following steps:

[0010] Step S1, obtaining a preoperative cardiac CT image to be predicted;

[0011] Step S2, extracting the imaging parameters of epicardial fat from the preoperative cardiac CT image to be predicted;

[0012] Step S3, according to the image group score Radscore of the image group parameters extracted in step S2, the image group score Radscore of the heart CT image to be predicted is calculated;The calculation formula of the image group score Radscore is:

[0013] Radscore=0.43713186×X1+(-0.07561088)×X2+0.46241184×X3+0.67098195×X4+(-0.22286110)×X5+0.11260365×X6+1.00656474;

[0014] Wherein, X1, X2, X3, X4, X5, X6 respectively represent the original image-morphology-maximum longitudinal two-dimensional diameter, three-dimensional 1 millimeter Laplace transform Gaussian filter-gray correlation matrix-homogeneity, three-dimensional 3 millimeter Laplace transform Gaussian filter-neighborhood gray difference matrix-roughness, high-low-low wavelet transform filter-voxel intensity distribution-kurtosis, high-low-low wavelet transform filter-gray run length matrix-gray difference, high-low-low wavelet transform filter-gray size region matrix-gray difference;

[0015] Step S4, according to the image group score Radscore, the risk probability Risk of left ventricular mass regression after transcatheter aortic valve replacement is calculated, and the Risk calculation formula is:

[0016] Risk=e (0.3674-1.4854×Radscore) / (1+e (0.3674-1.4854×Radscore) )。

[0017] Further, in step S2, the image group parameters of the heart CT image to be predicted are extracted, and the specific mode is:

[0018] The preoperative heart CT image to be predicted is input into Horos software, and the heart region of the heart CT image is outlined;The epicardial fat region in the heart region is extracted by inputting the outlined heart region into 3Dslicer software;Then the epicardial fat region is input into Pyradiomics component, and the image group parameters of the heart CT image to be predicted are extracted.

[0019] The beneficial effects of the present application are as follows:

[0020] In the present application, the heart CT image to be measured is obtained, and the EAT image group parameters are obtained according to the CT image, then the image group score Radscore is calculated by using the innovative algorithm, and finally the risk probability Risk of left ventricular mass regression is calculated by using the specific algorithm, so as to realize the effective prediction of the risk of left ventricular mass regression after transcatheter aortic valve replacement after left ventricular reverse remodeling, and the prediction accuracy of the risk value is high. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 is a flowchart of the present application;

[0022] Figure 2 is a risk prediction nomogram of poor regression of LVMi in the present application;

[0023] Figure 3 is an operating characteristic curve diagram of the risk prediction nomogram model of poor regression of LVMi in the present application;

[0024] Figure 4 is a calibration curve diagram of the risk prediction nomogram model of poor regression of LVMi in the present application;

[0025] Figure 5 is a decision curve diagram of the risk prediction nomogram model of poor regression of LVMi in the present application. DETAILED DESCRIPTION

[0026] 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 described clearly and completely below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application.

[0027] Therefore, all other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present application without creative labor fall within the scope of protection of the present application.

[0028] Embodiment 1

[0029] The present embodiment provides a method for predicting the risk of poor regression of left ventricular mass after transcatheter aortic valve replacement, which is used to solve the technical problem of effectively predicting the risk of poor regression of left ventricular mass after transcatheter aortic valve replacement.

[0030] It comprises the following steps:

[0031] Step S1, obtaining a preoperative cardiac CT image to be predicted;

[0032] Step S2, extracting extrapericardial fat imageomic parameters according to the preoperative cardiac CT image to be predicted;

[0033] The imageomic parameters of the cardiac CT image to be predicted are extracted in the following specific manner:

[0034] The preoperative heart CT image to be predicted is input into the Horos software, and the heart region of the heart CT image is outlined; the outlined heart region is input into the 3Dslicer software, and the epicardial fat region in the heart region is extracted; and then the epicardial fat region is input into the Pyradiomics component to extract the radiomics parameters of the preoperative heart CT image to be predicted.

[0035] Step S3, according to the radiomics parameters extracted in step S2, the radiomics score Radscore of the preoperative heart CT image to be predicted is calculated; wherein the calculation formula of the radiomics score Radscore is:

[0036] Radscore=0.43713186×X1+(-0.07561088)×X2+0.46241184×X3+0.67098195×X4+(-0.22286110)×X5+0.11260365×X6+1.00656474;

[0037] Wherein, X1, X2, X3, X4, X5, X6 respectively represent original image-morphology-maximum longitudinal two-dimensional diameter, three-dimensional 1 millimeter Laplace transform Gaussian filter-gray correlation matrix-homogeneity, three-dimensional 3 millimeter Laplace transform Gaussian filter-neighborhood gray difference matrix-roughness, high-low-low wavelet transform filter-voxel intensity distribution-kurtosis, high-low-low wavelet transform filter-gray run length matrix-gray difference, high-low-low wavelet transform filter-gray size region matrix-gray difference.

[0038] Step S4, according to the radiomics score Radscore, the risk probability Risk of poor regression of left ventricular mass after transcatheter aortic valve replacement is calculated, and the Risk calculation formula is:

[0039] Risk=e (0.3674-1.4854×Radscore) / (1+e (0.3674-1.4854×Radscore) )。

[0040] Test example

[0041] First, this test example collects the preoperative clinical and epicardial fat CT image data of 28 patients who receive transcatheter aortic valve replacement (TAVR). According to the percentage change (LVMi%) of the left ventricular mass index (LVMi) (i.e. [preoperative LVMi-postoperative LVMi 1 year] / preoperative LVMi), the patients are divided into two groups: patients with LVMi% ≥ 15% are classified as LVMi regression good, and patients with LVMi% < 15% are classified as LVMi regression poor.

[0042] Secondly, the collected 28 CT image data were sequentially input into the Horos software, and the Horos software outlined the heart region in the heart CT image. Then, the outlined heart region was input into the 3Dslicer software, and the 3Dslicer software extracted the EAT region (defined as the image region with CT value of -10 and -190 HU) in the heart region. Then, the EAT region was input into the Pyradiomics component, and 6 groups of image features were extracted from the epicardial fat data of each CT image by the Pyradiomics software package.

[0043] Then, the 6 groups of image features of each CT image data were substituted into the calculation of the radiomics score Radscore, and the risk probability of poor regression of LVMi was further calculated by the radiomics score Radscore. The corresponding Radscore, the predicted risk probability of poor regression of LVMi, and the true risk probability of 28 patients are shown in Table 1. The nomogram model for predicting the risk Risk of poor regression of LVMi based on the above principle is shown in Figure 2 .

[0044]

[0045] Table 1. Radscore, the predicted risk probability of poor regression of LVMi, and the true risk probability table

[0046] Finally, the operating characteristic curve, the calibration curve, and the decision curve were used to evaluate the accuracy and effectiveness of the risk Risk prediction model of poor regression of LVMi. The operating characteristic curve evaluation results are shown in Figure 3 , the area under the curve (AUC) value of the prediction result is 0.743, the sensitivity is 0.895, and the specificity is 0.667, indicating good accuracy.

[0047] The calibration curve evaluates the effectiveness of the prediction risk result, as shown in Figure 4 . The unreliability test coefficient (U) value of the calibration curve is 0.161, the area of the operating characteristic curve {C (ROC)} is 0.743, the slope is 0.811, and the brier index is 0.163, which shows good effectiveness and accuracy.

[0048] The decision curve analysis verifies the clinical benefit result of the prediction risk Risk, as shown in Figure 5 . The decision curve indicates that the use of nomogram to predict the risk probability of poor regression of LVMi after TAVR can provide clinical net benefit.

[0049] Example 2

[0050] The embodiment provides a left ventricular mass regression poor risk prediction system after transcatheter aortic valve replacement, which comprises:

[0051] An image acquisition module is configured to acquire a heart CT image to be predicted.

[0052] A parameter extraction module is configured to extract radiomics parameters of the heart CT image to be predicted according to the heart CT image to be predicted.

[0053] The radiomics parameters of the heart CT image to be predicted are extracted in the following manner:

[0054] The heart CT image to be predicted is input into Horos software, and a heart region of the heart CT image is outlined; the outlined heart region is input into 3Dslicer software, and an EAT region in the heart region is extracted; and the EAT region is input into Pyradiomics software, and the radiomics parameters of the heart CT image to be predicted are extracted.

[0055] An image score calculation module is configured to calculate a radiomics score Radscore of the heart CT image to be predicted according to the radiomics parameters extracted by the parameter extraction module; wherein the calculation formula of the radiomics score Radscore is:

[0056] Radscore=0.43713186×X1+(-0.07561088)×X2+0.46241184×X3+0.67098195×X4+(-0.22286110)×X5+0.11260365×X6+1.00656474;

[0057] Wherein, X1, X2, X3, X4, X5 and X6 represent original image-morphology-maximum longitudinal two-dimensional diameter, three-dimensional 1 millimeter Laplace transform Gaussian filter-grayscale correlation matrix-homogeneity, three-dimensional 3 millimeter Laplace transform Gaussian filter-neighborhood grayscale difference matrix-roughness, high-low-low wavelet transform filter-voxel intensity distribution-kurtosis, high-low-low wavelet transform filter-grayscale run length matrix-grayscale difference, and high-low-low wavelet transform filter-grayscale size region matrix-grayscale difference, respectively.

[0058] A risk probability calculation module is configured to calculate a risk probability Risk of inverse reconstruction poor according to the radiomics score Radscore, and the calculation formula of the risk probability Risk of inverse reconstruction poor is:

[0059] Risk=e (0.3674-1.4854×Radscore) / (1+e (0.3674-1.4854×Radscore) )。

[0060] Embodiment 3

[0061] A computer device comprising a memory and a processor, the memory storing a computer program, the computer program, when executed by the processor, causing the processor to perform the steps of the method for predicting the risk of poor left ventricular mass regression after transcatheter aortic valve replacement.

[0062] The computer device can be a desktop computer, a notebook computer, a palm computer, a cloud server, or the like. The computer device can interact with a user through a keyboard, a mouse, a remote controller, a touchpad, a voice control device, or the like.

[0063] The memory comprises at least one type of readable storage medium, such as a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., an SD or D interface display memory, or the like), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, or the like. In some embodiments, the memory can be an internal storage unit of the computer device, such as a hard disk or a memory of the computer device. In other embodiments, the memory can also be an external storage device of the computer device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, or the like. Of course, the memory can also include both the internal storage unit and the external storage device of the computer device. In this embodiment, the memory is commonly used to store an operating system and various application software installed in the computer device, such as program codes of the method for predicting the risk of poor left ventricular mass regression after transcatheter aortic valve replacement. In addition, the memory can also be used to temporarily store various data that have been output or will be output.

[0064] The processor can be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments. The processor is generally used to control the overall operation of the computer device. In this embodiment, the processor is used to run program codes or process data stored in the memory, such as program codes of the method for predicting the risk of poor left ventricular mass regression after transcatheter aortic valve replacement.

[0065] Embodiment 4

[0066] A computer readable storage medium stores a computer program, which, when executed by a processor, causes the processor to perform steps of a method for predicting risk of left ventricular mass regression failure after transcatheter aortic valve replacement.

[0067] The computer readable storage medium stores an interface display program, which can be executed by at least one processor to cause the at least one processor to perform steps of the method for predicting risk of left ventricular mass regression failure after transcatheter aortic valve replacement as described above.

[0068] Those skilled in the art can clearly understand from the description of the above embodiments that the above-mentioned embodiment methods can be implemented by means of software and a necessary general hardware platform, of course, can also be implemented by hardware, but in many cases, the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a plurality of instructions for causing a terminal device (which can be a mobile phone, computer, server or network device, etc.) to execute the transcatheter aortic valve replacement left ventricular mass regression failure risk prediction method described in the embodiments of the present application.

Claims

1. A method for predicting the risk of poor left ventricular quality regression after transcatheter aortic valve replacement, characterized in that, Includes the following steps: Step S1: Obtain the preoperative cardiac CT image to be predicted; Step S2: Extract the radiomics parameters of epicardial fat based on the preoperative cardiac CT images to be predicted. Step S3: Calculate the Radscore based on the radiomics parameters extracted in Step S2; the formula for calculating the Radiomics score (Radscore) is as follows: Radscore = 0.43713186 × X1 + (-0.07561088) × X2 + 0.46241184 × X3 + 0.67098195 × Where X1, X2, X3, X4, X5, and X6 represent the original image - morphology - maximum vertical two-dimensional diameter, the three-dimensional 1 mm Laplacian transform Gaussian filter - gray-level correlation matrix - uniformity, the three-dimensional 3 mm Laplacian transform Gaussian filter - neighborhood gray-level difference matrix - roughness, the high-low-low type wavelet transform filter - voxel intensity distribution - kurtosis, the high-low-low type wavelet transform filter - gray-level run length matrix - gray-level difference, and the high-low-low type wavelet transform filter - gray-level size region matrix - gray-level difference, respectively. Step S4: Calculate the risk of poor left ventricular quality regression (Risk) after transcatheter aortic valve replacement based on the Radscore (radiomics score). The formula for calculating Risk is: Risk=e (0.3674-1.4854×Radscore) / (1+e (0.3674-1.4854×Radscore) )。 2. The method for predicting the risk of poor left ventricular quality regression after transcatheter aortic valve replacement as described in claim 1, characterized in that, In step S2, the radiomics parameters of the cardiac CT image to be predicted are extracted, specifically as follows: The preoperative cardiac CT image to be predicted is input into Horos software to delineate the cardiac region of the cardiac CT image; the delineated cardiac region is input into 3Dslicer software to extract the epicardial fat region within the cardiac region; then the epicardial fat region is input into the Pyradiomics component to extract the radiomics parameters of the cardiac CT image to be predicted.

Images