Cervical lesion area auxiliary positioning method and device and storage medium

Abstract

The invention discloses an auxiliary positioning method and device for a cervical lesion area and a storage medium, and belongs to the technical field of image processing, and the method comprises the steps: obtaining a sample after cervical conization, carrying out the positioning and area division of the sample, and determining a scanning path; acquiring an OCT image of the sample after conization of the cervix uteri, and performing image preprocessing operation on the OCT image; for the preprocessed OCT image, the serial numbers of section images where suspected focus areas are located in the x direction, the y direction and the z direction are screened out from the B-Scan section images stacked in the x direction, the B-Scan section images stacked in the y direction and the B-Scan section images stacked in the z direction; performing suspected focus area sketching on the cross section image of the screening result serial number; and combining the suspected focus areas in the cross-section images in all directions to form three-dimensional volume data, and positioning the area numbers of the three-dimensional volume data in the area division of the sample. According to the invention, the acquired cervical data can be processed, the suspected focus position can be accurately positioned, and the omission ratio is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a method, a device and a storage medium for auxiliary positioning of a cervical lesion area. Background Art

[0002] Cervical lesions and cervical cancer are common types of diseases in gynecology, among which severe cervical cancer may threaten the patient's life. In current clinical practice, cervical conization and biopsy are important means to diagnose cervical lesions and cervical cancer. In this process, the doctor cuts a cone-shaped tissue sample from the patient's cervical os through cervical conization surgery and cuts it along its long axis to form a fan-shaped expansion. Among them, the tip of the cone corresponds to the cutting edge of the internal cervical os, and the bottom corresponds to the cutting edge of the external cervical os. For the convenience of subsequent observation and analysis, doctors usually further subdivide this fan-shaped sample into 12 parts, and the thickness of each tissue is roughly controlled at 2 to 3 mm. However, in the face of patients with cervical hypertrophy or excessively thick tissue, the number of sample subdivisions may need to be increased to 24 or 16 to ensure the comprehensiveness and accuracy of the diagnosis. Finally, these cone-cut samples will be stained and then placed under a microscope for observation to check for lesions.

[0003] Although cervical cone biopsy plays a key role in the diagnosis of cervical lesions, its "blind" nature has become a problem that cannot be ignored. Since it is impossible to accurately determine the specific location of cervical lesions before cone biopsy, doctors are faced with a dilemma when operating: increasing the number of cone biopsy samples and reducing the sample volume to reduce the missed diagnosis rate will lead to a significant increase in labor costs; while reducing the number of cone biopsies may increase the risk of missed diagnosis due to the inability to fully display the lesion site.

[0004] In recent years, optical coherence tomography (OCT) technology has been widely used in non-invasive detection of deep subcutaneous tissues due to its high resolution and deep imaging capabilities. For example, Chinese patent CN110110600B discloses a method, device and storage medium for identifying lesions in eye OCT images, which automatically identifies various types of lesions in eye OCT images by training classification models, effectively reducing time loss and costs. However, although OCT technology has shown great potential in lesion identification and localization, most of the current OCT-based lesion identification and localization methods rely on neural network models for classification. Such methods usually require a large amount of labeled clinical data for training, and when the amount of data is limited, their scope of application will be greatly limited.

[0005] In view of the "blind cutting" problem of cervical cone biopsy and the limitations of OCT technology in lesion identification, it is urgent to develop an OCT-based auxiliary positioning method for cervical lesion areas. This method aims to use the high-resolution imaging capability of OCT technology to perform non-invasive detection of cervical tissue and accurately locate the lesion area, thereby providing precise guidance for cervical cone biopsy. By achieving preliminary screening of the lesion area, it can effectively avoid the doctor's "blind cutting", reduce the risk of missed diagnosis, and improve the accuracy and efficiency of diagnosis of cervical lesions and cervical cancer. Summary of the invention

[0006] In view of the problems existing in the above-mentioned prior art, the present invention provides a method, device and storage medium for auxiliary positioning of cervical lesion area, which is intended to solve the problem of potential missed detection caused by the inability to accurately locate the location of cervical lesions during the existing cervical conization process due to too few cone cuts. To achieve the above-mentioned purpose, the present invention provides the following technical solutions:

[0007] A method for auxiliary positioning of cervical lesion area, comprising the following steps:

[0008] S1, obtain the sample after cervical conization, locate and divide the sample into regions, and determine the scanning path;

[0009] S2, obtaining an OCT image of the sample after cervical conization, and performing image preprocessing operations on the OCT image;

[0010] S3, for the preprocessed OCT images, the serial numbers of the cross-sectional images of the suspected lesion areas in the x-direction, y-direction and z-direction are respectively screened out from the B-Scan cross-sectional images stacked along the x-direction, the B-Scan cross-sectional images stacked along the y-direction and the B-Scan cross-sectional images stacked along the z-direction;

[0011] S4, outlining the suspected lesion area on the cross-sectional image of the screening result number;

[0012] S5, combining the suspected lesion areas in the cross-sectional images in each direction to form three-dimensional volume data, and locating the area number thereof in the area division of the sample.

[0013] Furthermore, the step S3 includes:

[0014] S31, loop through each B-Scan cross-sectional image stacked along the x direction, filter out the first n1 regions with the largest grayscale value in each image according to the statistical results of the cumulative distribution function of the statistical histogram, and use the connected domain marking method to filter out the binary image with the largest connected domain area; where 0 <n1<0.1;

[0015] S32, using the binary image result in S31, screening out the B-Scan image number of the suspected lesion area;

[0016] S33, repeating the above steps for the B-Scan cross-sectional images stacked along the y direction and the B-Scan cross-sectional images stacked along the z direction, to complete the sequence number screening of the cross-sectional images where the suspected lesion areas are located in three directions.

[0017] Furthermore, after step S32, the method further includes: further adjusting the B-Scan image sequence number of the suspected lesion area according to the continuity characteristics of the B-Scan image sequence number of the suspected lesion area.

[0018] Further, the step S4 includes: looping through each B-Scan cross-sectional image of the screening result sequence number, screening out the top n2 regions with the largest grayscale value in each image according to the statistical result of the cumulative distribution function of the statistical histogram, and screening out the binary image with the largest connected domain area by using the connected domain marking method, and taking the region as the suspected lesion region; wherein, 0 <n2<0.1。

[0019] Furthermore, the step S4 also includes: further screening of suspected lesion areas.

[0020] Furthermore, further screening of the suspected lesion area includes:

[0021] The suspected lesion areas are grouped, the suspected lesion areas of each group of B-Scan cross-sectional images are superimposed, and the template is obtained by threshold segmentation;

[0022] Calculate the area of ​​the overlap between the template and each B-Scan cross-sectional image of the suspected lesion area, and set the threshold value of the overlap area, Ares_2, where Ares_2 = T Area-2 ×M×N,T Area-2 The coefficient representing the area threshold, M and N are the height and width of the B-Scan cross-sectional image respectively;

[0023] The overlapping area is screened using the overlapping area threshold. If the overlapping area is larger than the threshold, the area is retained; if the overlapping area is smaller than the threshold, the suspected lesion area is replaced with a template.

[0024] Furthermore, it also includes multi-section auxiliary screening; judging the suspected lesion area in three directions, if the suspected lesion area is screened out in two or more of the three directions, it is determined to be a suspected lesion area.

[0025] Furthermore, the method of obtaining an OCT image of the sample after cervical conization includes: scanning the sample in blocks using an OCT device according to the results of the positioning and regional division of the sample to obtain the OCT images of the blocks; the image preprocessing includes one or more of image standardization, image denoising, image enhancement and image stitching.

[0026] A computer device comprises: a memory and a processor, wherein the memory and the processor are communicatively connected, the memory stores computer instructions, and the processor executes the above-mentioned method for auxiliary positioning of cervical lesion areas by executing the computer instructions.

[0027] A computer-readable storage medium stores computer instructions, wherein the computer instructions are used to enable a computer to execute the above-mentioned method for auxiliary positioning of cervical lesion areas.

[0028] The beneficial effects of the present invention are:

[0029] The present invention provides a method, device and storage medium for assisting in locating cervical lesion areas. By using an OCT device to perform a comprehensive scan of cervical samples, and combining an OCT three-dimensional image processing algorithm, the acquired cervical data is processed and analyzed to accurately locate the suspected lesion. The present invention can accurately guide the 12-point cone cutting of the cervix, that is, the knife is cut densely in the suspected lesion area and the knife is cut sparsely in the area without pathological indication, so as to reduce the workload of medical staff and reduce the missed diagnosis rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a flow chart of a method for auxiliary positioning of cervical lesion area provided by the present invention;

[0031] Figure 2 It is a schematic diagram of a cone-shaped sample cut from the cervical os provided in an embodiment of the present invention;

[0032] Figure 3 is a B-Scan cross-sectional image read in the embodiment of the present invention; wherein, Figure 3 (a) is a B-Scan cross-sectional original image read in an embodiment of the present invention; Figure 3 (b) Yes Figure 3 (a) Result image after preprocessing;

[0033] Figure 4 is the result image of step 3 in the embodiment of the present invention; wherein, Figure 4 (a) is a binary image drawn with the top 5% of the grayscale values ​​of a B-Scan cross-sectional image in an embodiment of the present invention; Figure 4 (b) is a result image after morphological operation is performed on the binary image in an embodiment of the present invention; Figure 4 (c) is a result image after connected domain screening is performed on the morphological operation image in an embodiment of the present invention; Figure 4 (d) is a result image after image grouping and superposition in an embodiment of the present invention; Figure 4 (e) is the result image of the image threshold segmentation in the embodiment of the present invention;

[0034] Figure 5 is the B-Scan image number selected in the embodiment of the present invention;

[0035] Figure 6 is the result image of step 4 in the embodiment of the present invention; wherein, Figure 6 (a) is the result image of binary image superposition according to an embodiment of the present invention; Figure 6 (b) is the image segmented after image superposition according to the embodiment of the present invention; Figure 6 (c) is the result image after image expansion according to the embodiment of the present invention; Figure 6 (d) is a schematic diagram of the B-Scan cross-sectional results of an embodiment of the present invention;

[0036] Figure 7 3D result schematic diagram of an embodiment of the present invention. DETAILED DESCRIPTION

[0037] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation methods, but the present invention is not limited to the following embodiments.

[0038] Embodiment 1:

[0039] See attached Figures 1 to 7 The present invention provides a method for auxiliary positioning of cervical lesion area, the process of which is as follows: Figure 1 As shown, in this embodiment, a cone-shaped sample is cut from the cervical opening as an example. The specific process is as follows:

[0040] S1, obtain the sample after cervical conization, locate and divide the sample into regions, and determine the scanning path;

[0041] Specifically, the cervical sample after cone sectioning is positioned, and the cervical cone section sample is divided into regions according to the size of the sample and the range of the OCT probe scanning field of view, and the final scanning path of the OCT probe is determined. Figure 2 As shown, in this embodiment, the cervical conization sample is divided into 5 areas, and the arrows are the scanning paths of the OCT probe.

[0042] S2, obtaining an OCT image of the sample after cervical conization, and performing image preprocessing operations on the OCT image;

[0043] Specifically, according to the positioning and regional division results of the cervical conization sample, the OCT device is used to scan the cervical sample in different regions to obtain the divided OCT cervical conization three-dimensional data; Figure 3 As shown, Figure 3 (a) is a B-scan cross-sectional image read in. Figure 3 (b) is the result image after preprocessing the B-scan cross-sectional image;

[0044] The preprocessing process of OCT image data includes image denoising and enhancement and three-dimensional image stitching. Specifically, the process of image denoising and enhancement is as follows:

[0045] Image data reading and standardization: read the three-dimensional scanning results of OCT, and loop to obtain its B-Scan section from the three-dimensional data, and perform standardization operations on it. The calculation formula is: Where I0(x,y) represents the input image, min(*) represents the minimum grayscale value of the image, max(*) represents the maximum grayscale value of the image, [*] represents the rounding function, I standard (x,y) represents the normalized result image.

[0046] Image denoising: Each two-dimensional B-Scan cross-sectional image is traversed in turn, and a horizontal coherent fringe template is generated by segmenting multiple B-Scan cross-sectional images to remove the horizontal coherent fringe noise in the original three-dimensional OCT image; the wavelet threshold denoising method is used to remove the speckle noise in the original three-dimensional OCT image.

[0047] Image enhancement: The denoised image can be enhanced by using a contrast-limited adaptive histogram equalization method. The image denoising and enhancement method can use the denoising and enhancement method disclosed in a publication number of CN117830138A entitled A denoising and enhancement method for Fourier domain optical coherence tomography.

[0048] Image stitching: The image stitching method may adopt the image stitching method disclosed in the publication number CN118229550A, entitled A method for automatic stitching of three-dimensional images, computer equipment and medium.

[0049] S3, for the preprocessed OCT images, the serial numbers of the cross-sectional images of the suspected lesion areas in the x-direction, y-direction and z-direction are respectively screened out from the B-Scan cross-sectional images stacked along the x-direction, the B-Scan cross-sectional images stacked along the y-direction and the B-Scan cross-sectional images stacked along the z-direction;

[0050] Specifically, step S3 includes:

[0051] S31, loop through each B-Scan cross-sectional image stacked along the x direction, filter out the first n1 regions with the largest grayscale value in each image according to the statistical results of the cumulative distribution function of the statistical histogram, and use the connected domain marking method to filter out the binary image with the largest connected domain area; wherein, preferably, 0 <n1<0.1。

[0052] In one embodiment of the present application, n1=0.05. Specifically:

[0053] Extract a B-Scan cross-sectional image and obtain the grayscale statistical histogram of the image;

[0054] Draw its cumulative distribution function (CDF) according to the grayscale statistical histogram;

[0055] According to the cumulative distribution function, the top 5% of pixels with the largest grayscale values ​​are screened out, and the image is binarized according to the screening results. The top 5% of pixels with the largest grayscale values ​​are set to 1, and the rest are set to 0 to obtain a binary image result; Figure 4 As shown in (a), Figure 4 (a) A binary image drawn for the top 5% of the grayscale values;

[0056] Using the morphological dilation operation method, the bright areas that may be connected are further connected, such as Figure 4 (b)

[0057] The binary image is filtered according to the area, and only the binary area with the largest area is retained as the result, such as Figure 4 (c) as shown.

[0058] S32, using the binary image result in S31, filter out the B-Scan image number of the suspected lesion area, specifically:

[0059] A set of B-Scan binary images with consecutive serial numbers and length N1 are grouped together and image superposition is performed, such as Figure 4 (d)

[0060] The superimposed image is binarized using the threshold segmentation method, where the threshold can be set to Set the part greater than the threshold to 1, and the part less than the threshold to 0, and obtain the segmentation result, as shown in 4(e);

[0061] Divide the area of ​​the region where the pixel value in the segmentation result is 1 by the area of ​​the region where the pixel value in the overlay result is greater than or equal to 1 to obtain the area ratio;

[0062] Set the threshold T according to the area ratio Area-1 , preferably, TArea-1 Set it to 0.25 to filter out B-Scan image numbers that are greater than the area ratio threshold, such as Figure 5 As shown, the horizontal axis is the B-Scan sequence number and the vertical axis is the area ratio.

[0063] In a certain embodiment of the present application, after step S32, the B-Scan image number of the suspected lesion area is further adjusted according to the continuity characteristics of the B-Scan image number of the suspected lesion area, specifically:

[0064] First, the B-Scan image number of the suspected lesion area is mapped to a one-dimensional array, and the one-dimensional array corresponding to the B-Scan number of the suspected lesion area is set to 1, and the one-dimensional array corresponding to the B-Scan number of the undiagnosed lesion area is set to 0;

[0065] Using the idea of ​​image logic closing operation, the one-dimensional array is closed to eliminate the discontinuous 0s in the B-Scan numbers of the continuous suspected lesion areas;

[0066] Count the lengths of the elements in the one-dimensional array and eliminate the elements whose lengths are less than the threshold T. N The B-Scan image number. Generally, T N =10.

[0067] S33, repeating the above steps for the B-Scan cross-sectional images stacked along the y direction and the B-Scan cross-sectional images stacked along the z direction, to complete the sequence number screening of the cross-sectional images where the suspected lesion areas are located in three directions.

[0068] S4, outlining the suspected lesion area on the cross-sectional image of the screening result number;

[0069] Specifically, step S4 includes:

[0070] Each B-Scan cross-sectional image of the screening result sequence number is traversed in a loop, and the top n2 regions with the largest grayscale value in each image are screened out according to the statistical results of the cumulative empirical distribution function of the statistical histogram, and the binary image with the largest connected domain area is screened out using the connected domain marking method, and the region is taken as the suspected lesion region; wherein, preferably, 0 <n2<0.1。

[0071] In a certain embodiment of the present application, n2=0.05. Specifically:

[0072] Extract a filtered B-Scan cross-sectional image and draw its grayscale statistical histogram;

[0073] Draw its cumulative distribution function based on the grayscale statistical histogram;

[0074] The cumulative distribution function is used to filter out the top 5% of pixels with the largest grayscale values, and the image is binarized according to the screening structure, the top 5% of pixels with the largest grayscale values ​​are set to 1, and the rest are set to 0 to obtain a binary image result;

[0075] The morphological opening operation is used to remove the stray bright spots in the image, and then the morphological dilation is used to connect the bright areas that may be connected;

[0076] Screening is performed based on the maximum value of the connected domain area, and only the binary region image with the largest area is retained, and this region is regarded as the suspected lesion area.

[0077] In one embodiment of the present application, step S4 further includes further screening of suspected lesion areas, specifically,

[0078] The suspected lesion areas of each group of B-Scans in step S4 are superimposed, and the superposition result is as follows: Figure 6 (a) and count the number of B-Scan images in each group N num ;

[0079] The result of superimposing each group of images is segmented using the threshold segmentation method, and the threshold is set to half the number of images in each group. The image threshold segmentation result is as follows: Figure 6 (b) as shown.

[0080] Perform image logical closing operation on the result of threshold segmentation to fill the empty areas in the binary area, such as Figure 6 As shown in (c), the result of the image logic operation is used as a template;

[0081] Calculate the area of ​​the overlap between the template and each B-Scan image of the suspected lesion, and set the threshold of the overlap area Area_2. Generally, Area_2 = T Area-2 ×M×N, where M and N are the height and width of the B-Scan image, respectively. Preferably, T Area-2 =0.1;

[0082] The overlapping area is screened using the area threshold. If the overlapping area is larger than the threshold, the area is retained; if the overlapping area is smaller than the threshold, the suspected lesion area is replaced with a template.

[0083] Finally, the boundary of the suspected lesion area is obtained and drawn with lines in the original image, such as Figure 6 (d) as shown.

[0084] Repeat the above steps to complete the positioning of the suspected lesion area in three directions.

[0085] In one embodiment of the present application, the auxiliary positioning method for cervical lesion area further includes multi-section auxiliary screening: judging the suspected lesion area in three directions, if two or more of the suspected lesion areas are screened out in the three directions, it is determined to be a suspected lesion area,

[0086] S5, combining the suspected lesion regions in the cross-sectional images in each direction to form three-dimensional volume data, and locating the region number thereof in the region division of the sample. Specifically, combining the suspected lesion regions in the cross-sectional images in each direction to form three-dimensional volume data, and locating the region number thereof in the region division of the sample according to the segmentation result of the three-dimensional volume data.

[0087] According to the region number of the abnormal region detection result, it can be densely cut; for the non-marked abnormal region, it can be sparsely cut, so as to further divide the cone sample. Finally, the cone-sectioned sample is stained and sliced ​​to make a pathological specimen, which is further observed under an optical microscope to obtain the final test result.

[0088] Figure 7 This is the positioning result of a lesion area of ​​a cervical sample obtained by the method of the present invention. The red area is the suspected lesion area marked. Among them, there are suspected lesion areas at 1 o'clock, 2 o'clock, 3 o'clock, 4 o'clock, 5 o'clock, 6 o'clock, 7 o'clock, 8 o'clock, 9 o'clock, 10 o'clock, 11 o'clock and 12 o'clock. The traditional pathological diagnosis results of the hospital are: high-grade squamous intraepithelial lesions involving glands are present at 8 o'clock, 9 o'clock, 10 o'clock and 11 o'clock of the cervical sample, and chronic inflammation is present at 1 o'clock, 2 o'clock, 3 o'clock, 4 o'clock, 5 o'clock, 6 o'clock, 7 o'clock and 12 o'clock. Therefore, the positioning results of the suspected lesion area of ​​the present application show significant consistency with the traditional pathological diagnosis results of the hospital.

[0089] Embodiment 2:

[0090] See attached Figures 1 to 7 . On the basis of the first embodiment, this embodiment further provides an electronic device, including a memory and a processor, wherein the memory and the processor are communicatively connected, the memory stores computer instructions, and the processor executes the cervical lesion area auxiliary positioning method in the above embodiment by executing the computer instructions. Specifically, all or part of the steps of the above method embodiment can be completed by hardware related to program instructions, and the computer program can be stored in a computer-readable storage medium. When the computer program is executed, it executes the steps of the above method embodiment; and the storage medium may include: mobile storage devices, read-only memories (ROM), magnetic disks or optical disks, and other media that can store program codes.

[0091] It is understandable that the above method of using OCT retrieval to determine suspected disease areas can be used not only for ex vivo cervical tissue, but also for pre-examination of other ex vivo tissues. This method can non-invasively detect lesion areas in various tissue samples through high-resolution imaging technology, providing doctors with more accurate diagnostic basis. Whether it is for cervical lesions or other types of tissue lesions, this method can significantly improve the accuracy and efficiency of lesion localization, reduce the risk of missed diagnosis, and thus improve the overall medical level. In addition, the widespread application of this technology can also help reduce medical costs and alleviate the economic burden on patients, and has important clinical significance and social value.

[0092] Embodiment three:

[0093] See attached Figures 1 to 7 On the basis of the second embodiment, this embodiment further provides a computer-readable storage medium, on which computer instructions are stored, and the computer instructions are used to enable a computer to execute the auxiliary positioning method for the cervical lesion area in the above embodiment.

[0094] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for auxiliary positioning of cervical lesion area, characterized in that: The following steps are involved: S1, obtain the sample after cervical conization, locate and divide the sample into regions, and determine the scanning path; S2, obtaining an OCT image of the sample after cervical conization, and performing image preprocessing operations on the OCT image; S3, for the preprocessed OCT images, the serial numbers of the cross-sectional images of the suspected lesion areas in the x-direction, y-direction and z-direction are respectively screened out from the B-Scan cross-sectional images stacked along the x-direction, the B-Scan cross-sectional images stacked along the y-direction and the B-Scan cross-sectional images stacked along the z-direction; S4, outlining the suspected lesion area on the cross-sectional image of the screening result number; S5, combining the suspected lesion areas in the cross-sectional images in each direction to form three-dimensional volume data, and locating the area number thereof in the area division of the sample.

2. The method for auxiliary positioning of cervical lesion area according to claim 1, characterized in that: The step S3 comprises: S31, loop through each B-Scan cross-sectional image stacked along the x direction, filter out the first n1 regions with the largest grayscale value in each image according to the statistical results of the cumulative distribution function of the statistical histogram, and use the connected domain marking method to filter out the binary image with the largest connected domain area; where 0 <n1<0.1; S32, using the binary image result in S31, screening out the B-Scan image number of the suspected lesion area; S33, repeating the above steps for the B-Scan cross-sectional images stacked along the y direction and the B-Scan cross-sectional images stacked along the z direction, to complete the sequence number screening of the cross-sectional images where the suspected lesion areas are located in three directions.

3. The method for auxiliary positioning of cervical lesion area according to claim 2, characterized in that: The step S32 further includes: further adjusting the B-Scan image sequence number of the suspected lesion area according to the continuity characteristics of the B-Scan image sequence number of the suspected lesion area.

4. The method for auxiliary positioning of cervical lesion area according to claim 1, characterized in that: The step S4 comprises: looping through each B-Scan cross-sectional image of the screening result sequence number, screening out the first n2 regions with the largest grayscale value in each image according to the statistical result of the cumulative distribution function of the statistical histogram, and screening out the binary image with the largest connected domain area by using the connected domain marking method, and taking the region as the suspected lesion region; wherein, 0 <n2<0.1。 5. The method for auxiliary positioning of cervical lesion area according to claim 4, characterized in that: The step S4 also includes: further screening of suspected lesion areas.

6. The method for auxiliary positioning of cervical lesion area according to claim 5, characterized in that: Further screening of the suspected lesion area includes: The suspected lesion areas are grouped, the suspected lesion areas of each group of B-Scan cross-sectional images are superimposed, and the template is obtained by threshold segmentation; Calculate the area of ​​the overlap between the template and each B-Scan cross-sectional image of the suspected lesion area, and set the threshold of the overlap area Area_2, where Area_2 = T Area-2 ×M×N,T Area_2 The coefficient representing the area threshold, M and N are the height and width of the B-Scan cross-sectional image respectively; The overlapping area is screened using the overlapping area threshold. If the overlapping area is larger than the threshold, the area is retained; if the overlapping area is smaller than the threshold, the suspected lesion area is replaced with a template.

7. The method for auxiliary positioning of cervical lesion area according to claim 5, characterized in that: It also includes multi-section auxiliary screening; the suspected lesion area in three directions is judged. If the suspected lesion area is screened out in two or more of the three directions, it is determined to be a suspected lesion area.

8. The method for auxiliary positioning of cervical lesion area according to claim 1, characterized in that: The method of obtaining the OCT image of the sample after cervical conization includes: scanning the sample in blocks using an OCT device according to the results of the positioning and regional division of the sample to obtain the OCT images of the blocks; the image preprocessing includes one or more of image standardization, image denoising, image enhancement and image stitching.

9. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected, the memory stores computer instructions, and the processor executes the method for auxiliary positioning of the cervical lesion area according to any one of claims 1 to 8 by executing the computer instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the cervical lesion area auxiliary positioning method according to any one of claims 1 to 8.

Images