A method and device for identifying vulnerable plaques

Abstract

The present application relates to the technical field of medical devices, and provides a method for identifying vulnerable plaques, including: obtaining N frames of light attenuation coefficient images corresponding to N temporally adjacent vascular OCT images, where N is a positive integer greater than 1; determining a target light attenuation coefficient image from the N frames of light attenuation coefficient images; determining a target IPA according to the IPA of M frames of light attenuation coefficient images adjacent to the target light attenuation coefficient image, the target IPA is positively correlated with the sum of the IPAs of the M frames of light attenuation coefficient images, M is an integer greater than 1, and M is less than or equal to N; when the value of the target IPA is greater than the IPA threshold, it is determined that the target light attenuation coefficient image contains a vulnerable plaque region; or, when the value of the target IPA is less than or equal to the IPA threshold, it is determined that the target light attenuation coefficient image does not contain a vulnerable plaque region. The above method can improve the discrimination of the region corresponding to the vulnerable plaque in the OCT image.

Description

Technical Field

[0001] This application relates to the technical field of medical devices, and particularly to a method and device for identifying vulnerable plaques. Background Art

[0002] As is well known, a vulnerable plaque refers to a plaque that is unstable and prone to thrombosis. Due to the characteristics of a very thin surface envelope, a large internal lipid content, and a large amount of inflammatory substances in vulnerable plaques, they are prone to rupture, bleeding, calcification, or thrombosis formation. Research shows that vulnerable plaques are closely related to the occurrence of cardiovascular diseases and are the main cause of diseases such as thrombosis, acute coronary syndrome, and coronary heart disease. Therefore, accurately judging the presence and severity of cardiovascular vulnerable plaques is of great significance in the prevention and diagnosis of cardiovascular diseases.

[0003] Currently, there are many methods for detecting cardiovascular vulnerable plaques. For example, intravascular ultrasound, optical coherence tomography (OCT) technology, etc. Medical staff need to analyze OCT images to determine whether there are vulnerable plaques in the cardiovascular system of patients. However, the lipids in the cardiovascular vulnerable plaques shown in existing OCT images are darker in color, resulting in a lower degree of differentiation between the vulnerable plaques and other surrounding tissues.

[0004] Therefore, how to improve the differentiation degree of the area corresponding to the vulnerable plaque in the OCT image is an urgent problem to be solved currently. Summary of the Invention

[0005] This application provides a method for identifying vulnerable plaques, which can improve the differentiation degree of the area corresponding to the vulnerable plaque in the OCT image.

[0006] In a first aspect, a method for identifying vulnerable plaques is provided, including: obtaining N frame light attenuation coefficient images corresponding to N temporally adjacent vascular OCT images, where N is a positive integer greater than 1; determining a target light attenuation coefficient image from the N frame light attenuation coefficient images; determining a target IPA according to the IPA of M frame light attenuation coefficient images adjacent to the target light attenuation coefficient image, the target IPA being positively correlated with the sum of the IPAs of the M frame light attenuation coefficient images, M being an integer greater than 1, and M being less than or equal to N; when the value of the target IPA is greater than the IPA threshold, determining that the target light attenuation coefficient image contains a vulnerable plaque area; or when the value of the target IPA is less than or equal to the IPA threshold, determining that the target light attenuation coefficient image does not contain a vulnerable plaque area.

[0007] The above method can be executed by a terminal device or a chip in the terminal device. Since the light attenuation coefficient of lipid plaques is relatively large, and vulnerable plaques usually contain a large lipid core, it is possible to analyze the distribution position of vulnerable plaques by obtaining the light attenuation coefficient image corresponding to the vascular OCT image. When a cardiovascular disease occurs, it usually presents as a continuous diseased segment. Therefore, the accuracy of determining vulnerable plaques by only analyzing the IPA of a single light attenuation coefficient image is relatively low. In this application, the target light attenuation coefficient image is the light attenuation coefficient image of the vascular region suspected of containing vulnerable plaques, and the IPA of the M light attenuation coefficient images adjacent to the target light attenuation coefficient image reflects the diseased conditions of a continuous region in the blood vessel. Therefore, the target IPA calculated based on the IPA of the M light attenuation coefficient images can improve the distinguishability of the region corresponding to the vulnerable plaque in the OCT image.

[0008] Optionally, determining the target light attenuation coefficient image from the N light attenuation coefficient images includes: generating a tiled image according to the N light attenuation coefficient images; displaying the tiled image; receiving indication information input by a user on the display interface of the tiled image; and determining the target light attenuation coefficient image from the N light attenuation coefficient images according to the indication information.

[0009] The tiled image is an image generated by combining multiple temporally consecutive light attenuation coefficient images, which can intuitively display the region where vulnerable plaques are distributed. The user can select the region of interest on the display interface of the tiled image based on experience and click on the region of interest to trigger the terminal device to generate indication information, thereby meeting the personalized needs of the user.

[0010] Optionally, generating the tiled image according to the N light attenuation coefficient images includes: determining N maximum light attenuation coefficient vectors corresponding to the N light attenuation coefficient images; and generating the tiled image according to the N maximum light attenuation coefficient vectors.

[0011] Optionally, the M light attenuation coefficient images are the 10 light attenuation coefficient images before the target light attenuation coefficient image and the 10 light attenuation coefficient images after the target light attenuation coefficient image.

[0012] Since a cardiovascular disease usually presents as a continuous diseased segment when it occurs, the user can determine the target IPA value of the target light attenuation coefficient image by analyzing the IPA values of a total of 20 light attenuation coefficient images before and after the target light attenuation coefficient image, so as to facilitate the user to accurately judge the position where vulnerable plaques exist in the target light attenuation coefficient image.

[0013] Optionally, the N temporally adjacent vascular OCT images are fused with the N light attenuation coefficient images to generate N fused images. The light attenuation distribution of the N temporally adjacent vascular OCT images can be visually observed from the N fused images.

[0014] Optionally, the fusing the N temporally adjacent vascular OCT images with the N light attenuation coefficient images to generate N fused images includes: staining the N light attenuation coefficient images to generate N stained light attenuation coefficient images; and fusing the N temporally adjacent vascular OCT images with the N stained light attenuation coefficient images to generate the N fused images.

[0015] The N light attenuation coefficient images can be stained according to the magnitudes of the light attenuation coefficients to generate N stained light attenuation coefficient images; then, the N temporally adjacent vascular OCT images are fused with the N stained light attenuation coefficient images to generate N fused images. The N fused images are color images, which can further improve the distinguishability of the regions corresponding to vulnerable plaques in the OCT images.

[0016] In a second aspect, a device for identifying vulnerable plaques is provided. The device includes a processor and a memory. The memory is configured to store a computer program, and the processor is configured to call and run the computer program from the memory, so that the device executes the method according to any one of the first aspect.

[0017] In a third aspect, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor is caused to execute the method according to any one of the first aspect. Description of the Drawings

[0018] To more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 It is a schematic flowchart of the method for identifying vulnerable plaques in an embodiment of the present invention;

[0020] Figure 2 It is a schematic diagram of an OCT image and a light attenuation coefficient image corresponding to the OCT image provided in an embodiment of the present invention;

[0021] Figure 3 It is a schematic diagram of the fusion of OCT and light attenuation coefficient images provided in an embodiment of the present invention;

[0022] Figure 4 is a schematic diagram of the carpet unfolding diagram provided by an embodiment of the present application;

[0023] Figure 5 is a schematic diagram of TCFA provided by an embodiment of the present application;

[0024] Figure 6 is a schematic diagram of FA provided by an embodiment of the present application;

[0025] Figure 7 is a schematic diagram of an OCT image with blood artifacts provided by an embodiment of the present application;

[0026] Figure 8 is a schematic diagram of a healthy blood vessel without lipids provided by an embodiment of the present application;

[0027] Figure 9 is a schematic diagram of an OCT thrombus provided by an embodiment of the present application;

[0028] Figure 10 is a schematic diagram of a guiding catheter image provided by an embodiment of the present application;

[0029] Figure 11 is a schematic diagram of the change in the value of AUC at different thresholds x provided by an embodiment of the present application;

[0030] Figure 12 is a schematic diagram of the ROC curve of the IPA for identifying the specificity and sensitivity of TCFA when the threshold x = 9.5 provided by an embodiment of the present application;

[0031] Figure 13 is a schematic diagram of the software display interface provided by an embodiment of the present application;

[0032] Figure 14 is a schematic diagram of the structure of a device for identifying vulnerable plaques provided by an embodiment of the present application. Detailed implementation manners

[0033] In the following description, for the purpose of illustration rather than limitation, specific details such as specific system structures and technologies are presented to thoroughly understand the embodiments of the present application. However, those skilled in the art should clearly understand that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, and methods are omitted to avoid unnecessary details from interfering with the description of the present application.

[0034] It should be understood that when used in the specification and the appended claims of the present application, the term "comprising" indicates the presence of the described features, wholes, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0035] It should also be understood that the term "and / or" used in the description of the present application specification and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0036] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used for differential description and cannot be understood as indicating or implying relative importance.

[0037] The reference to "one embodiment" or "some embodiments" etc. described in the present application specification means that in one or more embodiments of the present application, the specific features, structures or characteristics described in combination with that embodiment are included. Therefore, the statements "in one embodiment", "in some embodiments", "in other some embodiments", "in still other some embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "comprising", "including", "having" and their variants all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0038] The present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0039] Since vulnerable plaques are closely related to the occurrence of cardiovascular diseases and are the main causes of inducing diseases such as thrombosis, acute coronary syndrome, and coronary heart disease. And currently there are many methods for detecting cardiovascular vulnerable plaques. For example, intravascular ultrasound, optical coherence tomography (OCT) technology, etc. Usually, medical staff need to analyze OCT images to determine whether there are vulnerable plaques in the cardiovascular system of patients. However, the lipid color in the cardiovascular vulnerable plaques shown in the existing OCT images is darker, resulting in a lower distinguishability between the vulnerable plaques and other surrounding tissues, so that it is very difficult to identify the area where the vulnerable plaques exist on the OCT images. Therefore, how to improve the distinguishability of the corresponding area of the vulnerable plaques in the OCT images is an urgent problem to be solved currently.

[0040] A method for identifying vulnerable plaques provided by the present application can improve the distinguishability of the corresponding area of the vulnerable plaques in the OCT images. As Figure 1 shown, the method includes:

[0041] S101, obtaining N frame optical attenuation coefficient images corresponding to N temporally adjacent vascular OCT images, where N is a positive integer greater than 1.

[0042] Exemplarily, when an OCT machine scans a section of blood vessel in the human body, a set of OCT pullback data will be obtained. This set of OCT pullback data contains N frames of temporally adjacent blood vessel OCT images. For example, if the OCT machine scans a 4-mm-long blood vessel and acquires one frame of image for every 0.2-mm-long blood vessel, after scanning the 4-mm-long blood vessel, a set of OCT pullback data is obtained. This set of OCT pullback data contains N = 20 frames of blood vessel OCT images, and these 20 frames of images are arranged in the order of the scanning time.

[0043] Since the blood vessel OCT images (i.e., the original OCT images) acquired by the OCT machine are blood vessel OCT images in polar coordinates, therefore, a light attenuation model in polar coordinates can be used to calculate the light attenuation coefficient of each pixel point of a single-frame original OCT image, and use the light attenuation coefficient value corresponding to each pixel point to replace the value of each pixel point of the blood vessel OCT image, so as to obtain a single-frame light attenuation coefficient image corresponding to the single-frame original OCT image in polar coordinates. The method for calculating the N-frame light attenuation coefficient images corresponding to N frames of temporally adjacent blood vessel OCT images is the same as the method for calculating the single-frame light attenuation coefficient image corresponding to a single-frame original OCT image, and will not be elaborated here. The calculation formula of the above light attenuation model is as follows:

[0044]

[0045]

[0046]

[0047] Among them, I 0 is the scale factor, r represents the image depth, T(r) is the longitudinal point spread function, z 0 、z R 、z c and z w respectively represent the beam waist position, Rayleigh length, scanning center point and half-width of the roll-off function, and the values are 0, 3 mm, 0 and 10 μm respectively, u t is the light attenuation coefficient (i.e., the variable to be solved). Take the logarithm of both sides of formula (1), and then use the least squares method to calculate the light attenuation coefficient u t .

[0048] Taking a single-frame vascular OCT image acquired by an OCT machine as an example, it is illustrated how to obtain the corresponding single-frame optical attenuation coefficient image from the single-frame vascular OCT image, and how to perform fusion processing on the single-frame vascular OCT image and the corresponding single-frame optical attenuation coefficient image. Since the single-frame vascular OCT image acquired by the OCT machine is 642×500 (i.e., the vascular OCT image histogram), where 500 indicates that a total of 500 A-lines are scanned when the catheter performs a 360° scan in the blood vessel, and 642 indicates that 642 pixel points are scanned on each A-line. The optical attenuation coefficient of each pixel point of the single-frame vascular OCT image is calculated using the optical attenuation model to obtain the corresponding single-frame optical attenuation coefficient image of the vascular OCT image. This single-frame optical attenuation coefficient image is also 642×500 (i.e., the optical attenuation coefficient image histogram), where 500 indicates that a total of 500 A-lines are scanned when the catheter performs a 360° scan in the blood vessel, and 642 indicates that 642 pixel points are scanned on each A-line. The vascular OCT image histogram is converted into a circular vascular OCT image. Specifically, since a total of 500 A-lines are scanned when the catheter performs a 360° scan in the blood vessel, and one A-line is scanned every 0.72° (i.e., 360° divided by 500 equals 0.72°), then these 500 A-lines are arranged in a circle at an equal interval of 0.72° to obtain the circular vascular OCT image, as shown in Figure 2 (a), where 201 represents the guide wire, 202 represents the calibration cursor, and 203 represents the blood vessel wall. The method of converting the optical attenuation coefficient image histogram into a circular optical attenuation coefficient image is the same as the method of converting the vascular OCT image histogram into a circular vascular OCT image, which will not be elaborated here. This circular optical attenuation coefficient image is shown in Figure 2 (b), where 203 represents the blood vessel wall.

[0049] The staining interval is divided according to the magnitude of the optical attenuation coefficient, and the circular optical attenuation coefficient image is stained with a gradient color from blue to red to yellow. In addition, the magnitude of the optical attenuation coefficient represents the magnitude of the lipid plaque content. For example, blue is used to represent the optical attenuation coefficient value in the interval [0,3], red is used to represent the optical attenuation coefficient value in the interval [4,6], and yellow is used to represent the optical attenuation coefficient value in the interval [7,9]. The closer to blue, the smaller the optical attenuation coefficient and the smaller the lipid plaque content; the closer to yellow, the larger the optical attenuation coefficient and the larger the lipid plaque content. Therefore, the circular optical attenuation coefficient image can be stained according to the magnitude of the optical attenuation coefficient to obtain the stained circular optical attenuation coefficient image.

[0050] In order to better display the circular graph of the vascular OCT image and the corresponding circular graph of the optical attenuation coefficient image on the software interface, the bilinear interpolation algorithm is now used to perform linear interpolation on each pixel point of the circular graph of the vascular OCT image. Specifically, using the surrounding 4-neighborhood pixel points, linear interpolation is performed on the circular graph of the vascular OCT image in both the x and y directions to obtain the circular graph of the vascular OCT image after interpolation; the method of performing linear interpolation on the circular graph of the optical attenuation coefficient image corresponding to the circular graph of the vascular OCT image is similar to the method of performing linear interpolation on the circular graph of the vascular OCT image, which will not be elaborated here.

[0051] After linear interpolation, the circular graph of the vascular OCT image after interpolation and the corresponding circular graph of the optical attenuation coefficient image after interpolation are obtained. The circular graph of the vascular OCT image after interpolation and the corresponding circular graph of the optical attenuation coefficient image after interpolation are subjected to fusion processing, that is, first, each non-zero pixel value of the circular graph of the vascular OCT image after interpolation is counted, and then the optical attenuation coefficient value at the corresponding position of the corresponding circular graph of the optical attenuation coefficient image after interpolation is used to replace it, obtaining the fused OCT-optical attenuation coefficient image, as Figure 3 shown, where 203 represents the blood vessel wall.

[0052] The processing method for obtaining the corresponding N-frame optical attenuation coefficient images for N temporally adjacent vascular OCT images is the same as the processing method for obtaining the corresponding single-frame optical attenuation coefficient image for a single-frame vascular OCT image, which will not be elaborated here.

[0053] In addition, the method for fusing and generating N-frame fused images (i.e., N-frame OCT-optical attenuation coefficient images) for N temporally adjacent vascular OCT images and the corresponding N-frame optical attenuation coefficient images, for example, dyeing the N-frame optical attenuation coefficient images corresponding to N temporally adjacent vascular OCT images to generate N-frame dyed optical attenuation coefficient images, and fusing and processing the N-frame temporally adjacent vascular OCT images and the corresponding N-frame dyed optical attenuation coefficient images to generate N-frame fused images, is the same as the method for fusing and generating a single-frame fused image (i.e., a single-frame OCT-optical attenuation coefficient image) for a single-frame vascular OCT image and the corresponding single-frame optical attenuation coefficient image, which will not be elaborated here.

[0054] S102. Determine the target optical attenuation coefficient image from the N-frame optical attenuation coefficient images.

[0055] Exemplarily, according to the foregoing analysis, there are 500 A-lines in a single-frame optical attenuation coefficient image, and there are 642 optical attenuation coefficient values on each A-line. The maximum optical attenuation coefficient value on each A-line is calculated. There are 500 maximum optical attenuation coefficient values for the 500 A-lines. These 500 maximum optical attenuation coefficient values form a 1×500 maximum optical attenuation coefficient vector, that is, a single-frame optical attenuation coefficient image can obtain a 1×500 maximum optical attenuation coefficient vector. If a set of OCT pullback data contains 300 temporally adjacent vascular OCT images, 300 optical attenuation coefficient images will be obtained, and then 300 1×500 maximum optical attenuation coefficient vectors will be obtained. These 300 1×500 maximum optical attenuation coefficient vectors form a 300×500 maximum optical attenuation coefficient matrix, that is, the carpet plot, as Figure 4 shown, where 401 is the vulnerable plaque and 402 is the guide wire shadow.

[0056] Since the carpet plot is an image generated by combining multiple (i.e., N frames) temporally continuous optical attenuation coefficient images, and the N-frame optical attenuation coefficient images are obtained by using an optical attenuation model for N temporally adjacent vascular OCT images. This carpet plot can visually display the area where the vulnerable plaques are distributed, and the user can select the region of interest on the display interface of the carpet plot based on experience. For example, the user inputs an indication message on the display interface of the carpet plot. For instance, the user clicks on a certain position on the carpet plot (i.e., inputs an indication message for the region of interest on the carpet plot). At this time, the terminal device generates an indication message according to the position clicked by the user's mouse. This indication message is used to instruct the terminal device to determine the target optical attenuation coefficient image that the user wants to analyze from the N-frame optical attenuation coefficient images and display the target optical attenuation coefficient image on the interface for the user to further analyze, so as to meet the personalized needs of the user.

[0057] S103. Determine the target IPA according to the IPA of M temporally adjacent optical attenuation coefficient images of the target optical attenuation coefficient image. The target IPA is positively correlated with the sum of the IPAs of the M temporally adjacent optical attenuation coefficient images. M is an integer greater than 1, and M is less than or equal to N.

[0058] S104. When the value of the target IPA is greater than the IPA threshold, determine that the target optical attenuation coefficient image contains a vulnerable plaque region; or when the value of the target IPA is less than or equal to the IPA threshold, determine that the target optical attenuation coefficient image does not contain a vulnerable plaque region.

[0059] Exemplarily, the Index of plaque attenuation (IPA) is the proportion of the statistical light attenuation coefficient values greater than the threshold x, where the light attenuation coefficient represents the degree of light attenuation of different tissues during the OCT imaging process. From the above analysis, it can be seen that the carpet plot is a maximum light attenuation coefficient matrix, and each row of data in the carpet plot is a 1×500 maximum light attenuation coefficient vector (i.e., each row of data represents a frame of light attenuation coefficient image), where this maximum light attenuation coefficient vector has a total of 500 elements (i.e., 500 maximum light attenuation coefficient values μ t ) For the IPA value of a single frame of light attenuation coefficient image, it can be calculated using the following formula:

[0060]

[0061] where N(μ t >x) represents comparing these 500 elements with the threshold x respectively and counting the number of elements greater than the threshold x among these 500 elements. Since N total represents the total number of A-lines in the light attenuation coefficient image, and according to the foregoing analysis, 500 maximum light attenuation coefficient values μ t mean there are 500 A-lines, so N total takes the value of 500. For example, when N(μ t >x) is 300, N total takes the value of 500, and IPA is 600.

[0062] Exemplarily, for what value of the threshold x is more appropriate can be determined through experiments; in the case where the optimal threshold x is determined, the optimal IPA threshold can be determined through experiments, where the IPA threshold is used to determine whether the target light attenuation coefficient image contains a vulnerable plaque area. The experimental process of determining the optimal threshold x and determining the optimal IPA threshold in the case where this optimal threshold x is determined is as follows:

[0063] 1) Experimental data collection

[0064] The existing vascular OCT image data are preliminarily screened according to the following conditions (a) and (b). The vascular OCT image data that meet the experimental requirements should satisfy the following conditions:

[0065] a) Contain thin-cap fibroatheroma (TCFA) or fibroatheroma (FA), where the definition of TCFA is a plaque containing a lipid core and a fibrous cap thickness ≤ 65 μm, as Figure 5 shown, where 501 is TCFA; the definition of FA is a plaque containing a lipid core and a fibrous cap thickness > 65 μm, asFigure 6 As shown, where 601 is FA.

[0066] b) Since the above-mentioned TCFA belongs to vulnerable plaques (i.e., unstable plaques), when there is TCFA in the blood vessel, it indicates that the blood vessel has developed a lesion. For the OCT image data of the blood vessel that meets the condition of a), the lesion length in the OCT image data of this blood vessel should be ≥ 4 mm, that is, the number of frames containing the lesion should be ≥ 20 frames. The reason is that the OCT machine collects one frame of the blood vessel OCT image every 0.2 mm. If the OCT machine collects a 4-mm-long blood vessel in total, then 20 frames (i.e., 4 divided by 0.2) of blood vessel OCT images will be obtained finally.

[0067] Using the conditions of a) and b) to screen the OCT image data of the blood vessel, a total of 55 groups of OCT pullback data from 39 patients were collected. All the OCT image data of the blood vessel were collected by the P60 device and exported in dicom format. When the device exports the OCT image data of the blood vessel, it performs desensitization processing on the OCT image data of the blood vessel, such as deleting the personal information of the patient, etc.

[0068] Then, the 55 groups of OCT pullback data from 39 patients obtained by the preliminary screening are subjected to secondary screening according to the following conditions 1) to 4). The conditions for the secondary screening are as follows:

[0069] 1) For the image data with poor quality of the blood vessel OCT image caused by the artifact of insufficient blood clearance in the blood vessel OCT image, it should be excluded because it does not meet the experimental requirements, as Figure 7 shown. In the figure, 701 is the phenomenon of insufficient blood clearance. For example, in some patients, the blood clearance is not clean after surgery, resulting in poor quality of the blood vessel OCT image; for another example, the pullback operation of the OCT machine when collecting the blood vessel OCT image results in poor quality of the OCT image, etc.

[0070] 2) For the OCT image data of healthy blood vessels without lipids, it should be excluded because it does not meet the experimental requirements, as Figure 8 shown. In the figure, 801 is the healthy blood vessel wall. Since this experiment is to study how to determine the distribution area of lipid plaques from the optical attenuation coefficient image, and the healthy blood vessels without lipids are not the objects of this experiment.

[0071] 3) For the image data with poor quality of the blood vessel OCT image caused by the artifact of a large amount of thrombus in the blood vessel OCT image, it should be excluded because it does not meet the experimental requirements, as Figure 9 shown. In the figure, 901 is the thrombus. Since the blood vessel OCT image contains a large amount of thrombus artifacts, it will affect the judgment of lipid plaques in vulnerable plaques, so it does not meet the experimental requirements.

[0072] 4) For the image data containing the guiding catheter in the vascular OCT images, it also does not meet the experimental requirements. For example, Figure 10 As shown, 1001 in the figure is the guiding catheter. If the vascular OCT image contains a guiding catheter, the guiding catheter will affect the judgment of lipid plaques in vulnerable plaques, so it does not meet the experimental requirements.

[0073] After secondary screening of the OCT pull-back data of 55 groups from 39 patients, 19 segments of TCFA data and 88 segments of FA data from 39 groups of data of 31 patients were finally obtained. These 19 segments of TCFA data and 88 segments of FA data were used as the research objects of this experiment.

[0074] 2) Evaluation indicators of experimental results

[0075] Three cardiovascular experts from different top three hospitals were invited to use industry-recognized professional software, such as Labelme software, to jointly label the above 107 segments of experimental data (i.e., 19 segments of TCFA data and 88 segments of FA data), and the TCFA data and FA data marked by the experts were used as the gold standard for evaluating the experimental results. For example, the experts would label the 107 segments of experimental data one by one. Specifically, they would label whether each segment of data belonged to TCFA or FA, and finally the results marked by the experts were used as the gold standard for evaluating the experimental results.

[0076] 3) Experimental data processing and result analysis

[0077] Analysis was carried out using the Receiver Operating Characteristic Curve (ROC). The optimal threshold x was determined by calculating the area (Area Under Curve, AUC) enclosed by the ROC and the coordinate axes. The larger the value of the above AUC, the better the recognition ability of IPA for TCFA and FA. The ROC curve is a curve that reflects the relationship between sensitivity and specificity. Among them, sensitivity is the true positive, which refers to the proportion of the correct number among the identified positives to the total true positive number; specificity refers to the true negative, which refers to the proportion of the correct number among the identified negatives to the total true negative number; true positive means that the data contains TCFA, and true negative means that the data are all FA. And sensitivity and specificity are used to characterize the ability of IPA to recognize TCFA and FA. For example, there are a total of 107 segments of experimental data, among which, there are 19 segments of TCFA data and 88 segments of FA data. If 16 segments of TCFA data are identified and 80 segments of FA data are identified, then the sensitivity is 0.89 (i.e., 16 divided by 18), and the specificity is 0.91 (i.e., 80 divided by 88).

[0078] The experimental process for determining the above optimal threshold x is as follows: By fixing the IPA threshold and changing the threshold x, the recognition ability of IPA for TCFA and FA is studied. For example, when the IPA threshold is 200, an IPA value greater than 200 is judged as TCFA, and an IPA value less than or equal to 200 is judged as FA. As the threshold x changes, the change in the area of the ROC curve reflecting the ability of IPA to recognize TCFA and FA is shown in Table 1:

[0079] Table 1 Variation of the AUC value under different thresholds x

[0080] Threshold x AUC Threshold x AUC Threshold x AUC 6 0.3182 9 0.8702 12 0.8230 6.5 0.4773 9.5 0.8843 12.5 0.7775 7 0.14023 10 0.8666 13 0.7539 7.5 0.6932 10.5 0.8810 13.5 0.7404 8 0.7697 11 0.8657 14 0.6612 8.5 0.8541 11.5 0.81406 14.5 0.5520

[0081] From Table 1 and Figure 11 analysis shows that when the threshold x = 9.5, the value of AUC is the highest (i.e., 0.8843), indicating that when the threshold x = 9.5, the recognition ability of IPA for TCFA and FA is the best.

[0082] Under the condition that the threshold x = 9.5, the recognition performance of IPA for TCFA is further studied to determine the optimal IPA threshold. Specifically, under the condition that the threshold x is determined, the optimal IPA threshold is determined by changing the IPA threshold. Since the IPA threshold is used to judge whether the target optical attenuation coefficient image contains vulnerable plaque regions. Under the condition that the threshold x (for example, x = 9.5) is determined, when the IPA threshold is different, the results of judging TCFA and FA according to the IPA value of the target optical attenuation coefficient image are different. For example, there are a total of 107 segments of experimental data, among which, 19 segments are TCFA data and 88 segments are FA data. When the IPA threshold is 300, 12 segments of TCFA data and 70 segments of FA data are recognized; when the IPA threshold is 200, 17 segments of TCFA data and 85 segments of FA data are recognized. Thus, it can be seen that when the threshold x (for example, x = 9.5) is determined and the IPA threshold is 200, the recognition ability of this IPA threshold for TCFA and FA is the best.

[0083] Under the condition that the threshold x = 9.5, the specificity and sensitivity of the threshold IPA for recognizing TCFA are shown in Table 2. For example, there are a total of 19 segments of TCFA data, and 15 segments of TCFA data are recognized, that is, the sensitivity of TCFA is 0.79 (i.e., 15 divided by 19); there are still 3 segments of TCFA that are not recognized (i.e., TCFA is recognized as FA), that is, the specificity of TCFA is 0.16 (i.e., 3 divided by 19). It should be noted that the area containing TCFA is considered positive, and the area containing FA is considered negative.

[0084] Table 2 Results of the specificity and sensitivity of IPA for recognizing TCFA when the threshold x = 9.5

[0085] IPA value Specificity Sensitivity IPA value Specificity Sensitivity IPA value Specificity Sensitivity 0 0 1 70 0.6250 1 140 0.9091 0.4737 10 0.2273 1 80 0.7045 0.9474 150 0.9091 0.4737 20 0.3068 1 90 0.7614 0.8947 160 0.9091 0.4211 30 0.3977 1 100 0.8068 0.8947 170 0.9205 0.4211 40 0.4773 1 110 0.8523 0.7368 180 0.9205 0.3158 50 0.5341 1 120 0.8750 0.6842 190 0.9432 0.3158 60 0.14023 1 130 0.8977 0.5789 200 0.9545 0.2632

[0086] By analyzing Table 2 and Figure 12 performing the analysis, among which, Figure 12 the ROC curve reflects the relationship between 1 - specificity ("-" represents the subtraction symbol) and sensitivity changes. Among them, 1 - specificity is also called the false positive rate, and this false positive rate refers to the proportion of unrecognized true negatives (i.e., FA). When the threshold x = 9.5 and IPA = 100, the point on the ROC curve (the position indicated by the arrow, i.e., x = 1 - 0.8068 = 0.1932, y = 0.8947) is Figure 12 at the smallest distance from the position point in the upper left corner (i.e., x = 1 - 1 = 0, y = 1). This smallest distance indicates that both the sensitivity and specificity are relatively high at this time, and thus the comprehensive performance is better. Therefore, it is determined that when IPA = 100, the ability of IPA to detect the specificity and sensitivity of TCFA is the best.

[0087] Exemplarily, when a blood vessel undergoes a lesion (i.e., the blood vessel contains a lipid - core plaque), it usually presents as a continuous lesion segment. If the IPA values of consecutive frames (i.e., the IPA values of multiple optical attenuation coefficient images) are used to jointly determine the IPA value of the target frame (i.e., the IPA of the target optical attenuation coefficient image) to improve the accuracy of identifying vulnerable plaques (i.e., unstable plaques) with the IPA value of the target frame. For example, an OCT machine acquires a 10 - mm - long blood vessel. Since the OCT machine acquires one frame of blood vessel OCT image every 0.2 mm, after the OCT machine finishes acquiring this 10 - mm - long blood vessel, a total of N = 50 frames (i.e., 10 divided by 0.2) of blood vessel OCT images are finally obtained. The carpet plot composed of these 50 frames is as Figure 13 shown. In the figure, the area 1301 is used to display the cross - sectional view of the blood vessel lumen (i.e., a certain frame of blood vessel OCT image), the area 1302 is used to display the optical attenuation coefficient image (i.e., the optical attenuation coefficient image corresponding to a certain frame of blood vessel OCT image), the area 1303 is used to display the cross - sectional view of the blood vessel, the area 1304 is used to display the carpet plot, the area 1305 is used to display the L - axis diagram of the lumen, and 1306 is an OCT image indicator, which is used to indicate the area of interest to the user. When the user uses the OCT image indicator to Figure 13 mark a certain position of interest (i.e., the target optical attenuation coefficient image) on the carpet plot, if the IPA value of this target optical attenuation coefficient image needs to be calculated, the target IPA can be determined according to the IPA of M frames of optical attenuation coefficient images adjacent to this target optical attenuation coefficient image.

[0088] For example, using the images 2 mm before and after this target optical attenuation coefficient image, since the image spacing is 0.2 mm, there are a total of 20 frames of images before and after this target optical attenuation coefficient image (i.e., M = 20). Then, calculate the IPA of a total of 20 frames of images before and after this target optical attenuation coefficient image iValues, where i = 1, 2,..., 20. The target IPA (i.e., the IPA of the target optical attenuation coefficient image) is positively correlated with the sum of the IPAs of M optical attenuation coefficient images, where M is an integer greater than 1 and M is less than or equal to N. For example, in the above case, M = 20 and N = 50, obviously, M < N. Optionally, the sum of the IPAs of M optical attenuation coefficient images is used to determine the target IPA, that is, the target Alternatively, the average value of the IPAs of M optical attenuation coefficient images is used to determine the target IPA, that is, the target Of course, when M is an integer greater than 1 and M is less than or equal to N, the present application can also use the IPA values of the images 3 mm before and after the target optical attenuation coefficient image to jointly determine the IPA of the target optical attenuation coefficient image. The present application does not make any limitation on this.

[0089] Exemplarily, when the threshold x = 9.5 and the threshold IPA = 100, after determining the target IPA according to the IPA of M optical attenuation coefficient images, if the value of the target IPA is greater than the IPA threshold (i.e., 100), it is determined that the target optical attenuation coefficient image contains a vulnerable plaque region; or, when the value of the target IPA is less than or equal to the IPA threshold (i.e., 100), it is determined that the target optical attenuation coefficient image does not contain a vulnerable plaque region. For example, using the IPA values of the images 2 mm before and after the target optical attenuation coefficient image, that is, the IPA of 20 optical attenuation coefficient images, after determining the target IPA, for example, the target IPA is 150. Obviously, the target IPA (i.e., 150) is greater than the threshold 100, so it can be determined that the target optical attenuation coefficient image contains a vulnerable plaque region; if the target IPA is 80, obviously, the target IPA (i.e., 80) is less than the threshold 100, so it can be determined that the target optical attenuation coefficient image does not contain a vulnerable plaque region.

[0090] Figure 14 The figure shows a schematic structural diagram of a device for identifying vulnerable plaques provided by the present application. Figure 14 The dotted lines in... indicate that the unit or the module is optional. The device 1400 can be used to implement the method described in the above method embodiments. The device 1400 can be a terminal device, a server, or a chip.

[0091] The device 1400 includes one or more processors 1401, and the one or more processors 1401 can support the device 1400 to implement Figure 1The method in the corresponding method embodiment. The processor 1401 can be a general-purpose processor or a dedicated processor. For example, the processor 1401 can be a central processing unit (CPU). The CPU can be used to control the device 1400, execute software programs, and process the data of software programs. The device 1400 can also include a communication unit 1405 for implementing signal input (reception) and output (transmission).

[0092] For example, the device 1400 can be a chip, and the communication unit 1405 can be the input and / or output circuit of the chip, or the communication unit 1405 can be the communication interface of the chip, and the chip can be a component of the terminal device.

[0093] For another example, the device 1400 can be a terminal device, and the communication unit 1405 can be the transceiver of the terminal device, or the communication unit 1405 can be the transceiver circuit of the terminal device.

[0094] The device 1400 can include one or more memories 1402, on which there is a program 1404. The program 1404 can be run by the processor 1401 to generate instructions 1403, so that the processor 1401 executes the method described in the above method embodiment according to the instructions 1403. Optionally, data (such as the ID of the chip to be tested) can also be stored in the memory 1402. Optionally, the processor 1401 can also read the data stored in the memory 1402. The data can be stored at the same storage address as the program 1404, or the data can be stored at a different storage address from the program 1404.

[0095] The processor 1401 and the memory 1402 can be set separately or integrated together. For example, they can be integrated on a system on chip (SOC) of the terminal device.

[0096] The specific manner in which the processor 1401 executes the method for identifying vulnerable plaques can be referred to the relevant description in the method embodiment.

[0097] It should be understood that each step of the above method embodiment can be completed by a logic circuit in hardware form or an instruction in software form in the processor 1401. The processor 1401 can be a CPU, a digital signal processor (DSP), a field programmable gate array (FPGA), or other programmable logic devices. For example, discrete gate, transistor logic devices, or discrete hardware components.

[0098] The present application also provides a computer program product, which, when executed by a processor 1401, implements the method described in any of the method embodiments of the present application.

[0099] This computer program product can be stored in a memory 1402, for example, it is a program 1404. After processes such as preprocessing, compilation, assembly, and linking, the program 1404 is finally converted into an executable target file that can be executed by the processor 1401.

[0100] The present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by the computer, it implements the method described in any of the method embodiments of the present application. This computer program can be a high-level language program or an executable target program.

[0101] This computer-readable storage medium is, for example, the memory 1402. The memory 1402 can be a volatile memory or a non-volatile memory, or the memory 1402 can include both a volatile memory and a non-volatile memory at the same time. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchlink dynamic random access memory (SLDRAM), and direct rambus random access memory (DRRAM).

[0102] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes and the technical effects generated by the above-described devices and equipment can refer to the corresponding processes and technical effects in the foregoing method embodiments, and will not be elaborated herein.

[0103] In several embodiments provided by this application, the disclosed systems, devices, and methods can be implemented in other ways. For example, some features of the method embodiments described above can be ignored or not executed. The device embodiments described above are merely illustrative. The division of units is only a logical function division. In actual implementation, there may be other division methods. Multiple units or components can be combined or integrated into another system. In addition, the coupling between units or the coupling between each component can be direct coupling or indirect coupling. The above-mentioned coupling includes electrical, mechanical, or other forms of connection.

[0104] The above-described embodiments are only used to illustrate the technical solutions of this application, rather than to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features. These modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included in the protection scope of this application.

Claims

1. A method for identifying vulnerable plaques, characterized in that, the method comprises: obtaining N frames of optical attenuation coefficient images corresponding to N temporally adjacent vascular OCT images, where N is a positive integer greater than 1; generating a carpet plot based on the N frames of optical attenuation coefficient images; displaying the carpet plot; receiving indication information input by a user on the display interface of the carpet plot; determining a target optical attenuation coefficient image from the N frames of optical attenuation coefficient images according to the indication information; determining a target IPA according to the IPA of M frames of optical attenuation coefficient images adjacent to the target optical attenuation coefficient image, the target IPA being positively correlated with the sum of the IPA of the M frames of optical attenuation coefficient images, M being an integer greater than 1, and M being less than or equal to N; when the value of the target IPA is greater than the IPA threshold, determining that the target optical attenuation coefficient image contains a vulnerable plaque region; or, when the value of the target IPA is less than or equal to the IPA threshold, determining that the target optical attenuation coefficient image does not contain a vulnerable plaque region.

2. The method according to claim 1, characterized in that, generating the carpet plot based on the N frames of optical attenuation coefficient images includes: determining N maximum optical attenuation coefficient vectors corresponding to the N frames of optical attenuation coefficient images; generating the carpet plot according to the N maximum optical attenuation coefficient vectors.

3. The method according to claim 1 or 2, characterized in that, the M frames of optical attenuation coefficient images are 10 frames of optical attenuation coefficient images before the target optical attenuation coefficient image and 10 frames of optical attenuation coefficient images after the target optical attenuation coefficient image.

4. The method according to claim 1 or 2, characterized in that, further comprising: performing a fusion process on the N temporally adjacent vascular OCT images and the N frames of optical attenuation coefficient images to generate N frames of fused images.

5. The method according to claim 4, characterized in that, performing the fusion process on the N temporally adjacent vascular OCT images and the N frames of optical attenuation coefficient images to generate N frames of fused images includes: staining the N frames of optical attenuation coefficient images to generate N frames of stained optical attenuation coefficient images; performing a fusion process on the N temporally adjacent vascular OCT images and the N frames of stained optical attenuation coefficient images to generate the N frames of fused images.

6. An apparatus for identifying vulnerable plaques, characterized in that, the apparatus comprises a processor and a memory, the memory is used for storing a computer program, and the processor is used for calling and running the computer program from the memory, so that the apparatus executes the method according to any one of claims 1 to 5.

7. A computer-readable storage medium, characterized in that, the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor executes the method according to any one of claims 1 to 5.

Images