Method for Measuring Multi-Point Thickness of Submucosa for Endoscopic Submucosal Dissection

Abstract

This application discloses a multi-point submucosal thickness measurement method for endoscopic submucosal dissection (ESD). The method comprises the following steps: image acquisition: obtaining a pathological image obtained by endoscopic submucosal dissection; image preprocessing: enhancing the brightness and highlight contrast of the pathological image and smoothing the pathological image using Gaussian filtering; image feature extraction: extracting texture features from the pathological image using the local binary pattern (LBP) algorithm and generating a vertical coordinate feature map of the same size as the pathological image; feature fusion: superimposing the denoised image, texture feature image, and vertical coordinate feature map obtained from the preprocessing; submucosal pathological tissue segmentation: training the superimposed image using a UNet neural network for image segmentation; multi-point thickness calculation; and quality assessment. This application measures submucosal thickness using pathological tissue images, facilitating the assessment of ESD efficacy and surgical quality control.

Description

Technical Field

[0001] The present application relates to the technical field of endoscopic submucosal dissection (ESD), and in particular to a method for measuring the multi-point thickness of the submucosal layer used in endoscopic submucosal dissection. Background Art

[0002] Gastrointestinal diseases are common and recurring health issues worldwide, including but not limited to gastritis, gastric ulcers, colitis, Crohn's disease, gastrointestinal bleeding, and intestinal tumors. These diseases not only have a high incidence but also complex and diverse symptoms, severely impacting patients' health and quality of life. With the aging population and changing lifestyles, the incidence of gastrointestinal diseases is on the rise, severely impacting people's quality of life and health. Therefore, accurate diagnosis and effective treatment of gastrointestinal diseases are particularly important.

[0003] Medical endoscopes play an important role in diagnosing gastrointestinal diseases. Traditional endoscopic examinations usually require doctors to insert a long cable into the patient's body to image key areas and provide diagnostic and treatment plans. However, this wired endoscope may cause discomfort and pain to the patient due to its invasive nature. In addition, the use of such endoscopes also carries the risk of potential complications such as infection, perforation, and tearing. Capsule Endoscopy (CE) is an innovative medical device used in the field of modern gastroenterology diagnosis and treatment. It is usually composed of a miniature camera, a housing, a lighting system, a control circuit, a wireless communication module, and a battery. It examines the digestive tract in a painless, non-invasive manner. However, in the process of realizing this application, the applicant found that the current capsule endoscope detection method has at least the following technical problems:

[0004] 1. During each test, the wireless capsule endoscope captures more than 60,000 images. The large number of images collected by the capsule endoscope makes manual testing time-consuming and labor-intensive, increasing the workload of clinicians.

[0005] 2. Due to problems such as uneven illumination and blurred focus in capsule endoscopy images, feature extraction is difficult. Existing detection methods are unable to accurately identify irregular lesions in complex backgrounds and lack dedicated detection.

[0006] 3. Existing models exhibit insufficient feature capture capabilities when detecting small lesions, leading to an increased risk of missed and misdiagnosed lesions, especially in early diagnosis, where critical lesions may be missed.

[0007] 4. The deep learning models used in existing technologies lack generalization capabilities in specific scenarios and struggle to cope with a wide variety of gastrointestinal image types. Furthermore, although recent studies have proposed improvements, such as enriching dataset diversity through data augmentation techniques or extracting the fractal dimension of images by introducing differential box counting to improve detection performance, these methods still fail to effectively address the issues of insufficient feature extraction and limited ability to detect small lesions in capsule endoscopy images. Furthermore, these studies also lack dataset balance and model generalization capabilities, failing to fully meet clinical needs.

[0008] Therefore, the above problems lead to low efficiency and low detection accuracy of capsule endoscopy and the risk of missed diagnosis and misdiagnosis, thus limiting the effective application of capsule endoscopy in clinical diagnosis. Summary of the Invention

[0009] The embodiment of the present application provides a multi-point thickness measurement method for the submucosal layer for endoscopic submucosal dissection, which solves the current problems of low detection efficiency, low detection accuracy and the risk of missed diagnosis and misdiagnosis of capsule endoscopes.

[0010] In order to solve the above technical problems, this application is implemented as follows:

[0011] A method for measuring the thickness of the submucosal layer at multiple points for endoscopic submucosal dissection is provided, comprising the following steps:

[0012] Image acquisition: Acquisition of pathological images of the cuts made by endoscopic submucosal dissection;

[0013] Image preprocessing: Improve the brightness and highlight contrast of pathological images, and smooth the pathological images through Gaussian filtering;

[0014] Image feature extraction: extracting texture features of the pathological image using the local binary pattern (LBP) algorithm and generating a vertical coordinate feature map with the same size as the pathological image;

[0015] Feature fusion: superimpose the denoised image, texture feature image and ordinate feature map obtained by preprocessing to form a superimposed image for segmentation;

[0016] Submucosal pathological tissue segmentation: The UNet neural network is used to train the superimposed images for image segmentation tasks to extract the submucosal layer;

[0017] Multi-point thickness calculation: The skeleton of the submucosal layer is extracted using a skeletonization method. This skeleton is used as the extension direction of the submucosal layer at that location, and the thickness of the submucosal layer at that location is calculated using a vertical line. This allows for multi-point thickness measurement of the submucosal layer during endoscopic submucosal dissection.

[0018] Quality assessment: The image was divided into equal parts vertically by manually selecting the endpoints of the line and performing an equal division operation to assess the surgical quality.

[0019] In one embodiment, in the step of enhancing the brightness of the pathological image:

[0020] Adjust the background brightness of the pathological image, perform brightness adjustment on the V channel of the image in the HSV color space, and retain the hue (H) and saturation (S) channels to avoid image color distortion. The clipping histogram is defined as follows:

[0021] , represents the original histogram, , clipping threshold Defined as 180.

[0022] In one embodiment, in the step of enhancing the contrast of the bright part of the pathological image:

[0023] Gamma transform is used to perform nonlinear transformation on the image pixel values to improve the contrast of the bright part of the pathological image. The formula of gamma transform is: , A is the transformation coefficient, which is used to restore the transformed image to the original image bit space, γ is used as the exponent of the exponential calculation, and when γ is greater than 1, the image contrast in high brightness areas can be improved.

[0024] In one embodiment, in the step of smoothing the pathological image by Gaussian filtering:

[0025] The pathological image is processed by applying Gaussian filtering, using a standard deviation value to smooth the image and reduce the interference of noise and details. The Gaussian filtering formula is:

[0026] , G(x,y) is the image value after Gaussian filtering, x and y are the image pixel coordinates, and σ is the standard deviation of the Gaussian distribution.

[0027] In one embodiment, in the step of extracting texture features of the pathological image:

[0028] Converting color images of pathological images into grayscale images;

[0029] The LBP algorithm is used to process grayscale images. The LBP algorithm generates a texture feature map by comparing the grayscale values of each pixel with its neighboring pixels.

[0030] In one embodiment, the model architecture of the UNet neural network includes:

[0031] Convolution module: It includes two convolution operations, each of which is followed by batch normalization and ReLU activation function processing;

[0032] Downsampling path: Reduce the feature map resolution and increase the number of channels through convolution modules and maximum pooling operations;

[0033] Bottleneck layer: further processes and compresses features through convolution modules to obtain the most representative high-dimensional feature representation;

[0034] Feature splicing: The UNet model splices the features in the downsampling process onto the upsampled features according to the feature hierarchy for subsequent feature upsampling.

[0035] Upsampling path: During the upsampling process, deconvolution operations are used to gradually restore the resolution of the feature map. At the same time, the upsampled features are fused with the corresponding downsampled layer features through feature concatenation, and then the features are integrated and optimized through the convolution module.

[0036] Output layer: A convolutional layer is used to map the features to the number of output channels. This method is a two-class image segmentation method with an output channel of 1. The Sigmoid activation function is used to limit the output value between 0 and 1 to obtain the probability that each pixel belongs to the target category.

[0037] In one embodiment, the UNet neural network model architecture further includes:

[0038] Position attention mechanism: used to encode and weight the position information of the feature map before feature fusion or jump connection, so as to improve the model's sensitivity to position features and improve the model's segmentation accuracy.

[0039] In one embodiment, in the step of weighting the position information of the feature map:

[0040] Perform a linear transformation on the input feature map to obtain the query vector Q, key vector K and value vector V;

[0041] Calculate the dot product of Q and K through matrix multiplication, and then perform scaling and Softmax operations to obtain the attention weight matrix A;

[0042] Multiply the attention weight matrix A with the value vector V to obtain the weighted feature representation to enhance the model's ability to recognize the region of interest.

[0043] In one embodiment, in the step of using a UNet neural network to train the superimposed image for an image segmentation task:

[0044] Model training: The error loss is calculated using a combination of Dice parameters and mean square error, and the stochastic gradient descent algorithm is used to optimize the model parameters to adapt to image segmentation under the working conditions of 5-channel image feature input;

[0045] Processing segmented images: For false positive points and false negative points in the segmentation results, they are corrected by using the connected area filtering method, the holes of false negative points are closed by using hole filling combined with closing operations, and the edges of false negative points are processed by edge smoothing operations to ensure the accuracy of the segmentation results.

[0046] In one embodiment, in the step of closing the holes of false negative points of the segmented image using hole filling combined with a closing operation:

[0047] Determine the hole area: traverse each negative connected area in the segmented image, and for each negative connected area, check the pixels in the outer 8 connected areas. If all pixels are positive, mark the negative connected area as a hole area;

[0048] Hole filling: Modify all negative pixel values in each marked hole area to positive pixel values.

[0049] In the embodiments of this application, by preprocessing the surgical dissection tissue pathology images, and then using multiple feature extraction methods for feature enhancement, a convolutional neural network is used to segment the submucosal layer. At the same time, methods such as closing operations and connected region filtering are used to remove erroneous results. Combined with the backbone direction positioning method, multi-point thickness measurement of the submucosal layer of pathological tissue is finally achieved. In this way, the present application uses pathological tissue images to measure the thickness of the submucosal layer, which helps to determine the effectiveness of endoscopic submucosal dissection (ESD) and control the surgical quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0051] Figure 1 is a flowchart of the steps of the method for measuring the multi-point thickness of the submucosal layer for endoscopic submucosal dissection of the present application;

[0052] Figure 2 This is the gamma transform function diagram when γ is 5 in this application;

[0053] Figure 3 and Figure 4 is a schematic diagram of the gastric wall submucosal specimen image before and after gamma transformation of the present application;

[0054] Figure 5and Figure 6 Schematic diagram of the gastric wall submucosal specimen image before and after Gaussian filtering processing in the present application;

[0055] Figure 7 This is a schematic diagram of the LBP feature extraction effect of this application;

[0056] Figure 8 It is the image longitudinal position information feature map of this application;

[0057] Figure 9 is a schematic diagram of the superimposed images of this application. DETAILED DESCRIPTION

[0058] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are part of the embodiments of this application, not all of them. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0059] See also Figure 1 , which is a flowchart of the steps of the multi-point thickness measurement method of the submucosal layer for endoscopic submucosal dissection of the present application; as shown in the figure, the multi-point thickness measurement method of the submucosal layer for endoscopic submucosal dissection includes the following steps S1 to S7. Among them:

[0060] Step S1, image acquisition: Acquire pathological images cut by endoscopic submucosal dissection. Pathological images are obtained from the cut pathological tissue obtained by endoscopic submucosal dissection, provided by hospital experts, and uploaded to computers and storage devices. The collected images are processed to obtain a series of images. , the pathological sample data is expressed as .

[0061] Step S2, image preprocessing: enhance the brightness and highlight contrast of the pathological image, and smooth the pathological image through Gaussian filtering. The pathological images collected in the experiment may have uneven brightness, unclear edge information, etc. The first step is to use color and brightness screening to enhance the brightness of the image background, and set the image background to the highest brightness according to the color brightness threshold. Specifically, since the shooting environment of the pathological tissue may have poor lighting conditions, the background brightness of the pathological image is first adjusted, and the brightness of the V channel of the image is adjusted in the HSV color space, retaining the hue (H) and saturation (S) channels to avoid image color distortion. This method can avoid over-enhancing the brightness of the image, and the cropped histogram is defined as follows: , represents the original histogram, , clipping threshold Defined as 180.

[0062] As mentioned above, the second step is to correct the image brightness and improve the contrast of the bright parts based on the gamma correction method. Specifically, since the image brightness of the pathological tissue and background area in the gastric wall submucosal specimen image is high, its histogram is distributed in the high brightness area, and the contrast between the pathological tissue and the background is poor. Therefore, the gamma transform is used to perform a nonlinear transformation on the image pixel values to improve the contrast of the bright parts of the pathological image. The formula of the gamma transform is: The principle is to perform exponential calculation on the input image to achieve nonlinear transformation. Among them, A is the transformation coefficient, which is used to restore the transformed image to the original image bit space. γ is the exponent of the exponential calculation. When γ is greater than 1, the image contrast in high-brightness areas can be improved. The image is processed by applying gamma transformation, and the gamma value is selected as 5 to enhance the contrast of the bright area. The transformation function image is as follows Figure 2 This processing method effectively improves the details of the highlight area in the submucosal specimen image, making the important information more prominent. Since the shooting scene of the gastric submucosal specimen image is bright, the gamma transform method is used to enhance the contrast of the bright part. The effect is as follows Figure 3 and Figure 4 shown.

[0063] The third step is to remove image noise through Gaussian filtering and remove redundant features of the image. By applying Gaussian filtering to process the pathological image, a standard deviation value σ is used to smooth the image to reduce the interference of noise and details. This processing method can effectively suppress random noise in the submucosal specimen image while maintaining the integrity of important structures such as edges. The Gaussian filter formula is:

[0064] , G(x,y) is the image value after Gaussian filtering, x and y are the image pixel coordinates, and σ is the standard deviation of the Gaussian distribution.

[0065]

[0066] Figure (4) Gaussian filter sliding window

[0067] By selecting the appropriate σ value, Gaussian filtering can effectively preserve key edge information while smoothing the image, making the features of the submucosal specimen image clearer. In this digital image processing task, a convolution kernel of size 3×3 is used to perform Gaussian filtering on the discrete digital image to reduce noise while maintaining key details. The effect is as follows Figure 5 and Figure 6As shown in Figure 2, Gaussian filtering removes image noise to a certain extent and optimizes the holes and noise in the pathological image of the gastric wall submucosal specimen.

[0068] Step S3, image feature extraction: Extract the texture features of the pathological image using the local binary pattern (LBP) algorithm, and generate a vertical coordinate feature map of the same size as the pathological image. This embodiment uses an efficient image texture feature extraction method, which first converts the input color image into a grayscale image. In other words, the color image of the pathological image is converted into a grayscale image, thereby reducing the complexity of the calculation and highlighting the brightness information of the image. Subsequently, the local binary pattern (LBP) algorithm is used to process the grayscale image. The LBP algorithm generates a texture feature map by comparing the grayscale values of each pixel with those of its neighboring pixels. The specific principle is to compare the brightness of the pixel point with the surrounding pixels within a certain range. If it is higher than the comparison point, the feature is recorded as 1, otherwise it is 0. A decimal value is generated based on this binary sequence and used as the texture feature of the point. In this embodiment, the number of neighborhood pixels is set to 8, the neighborhood radius is set to 5, and the decimal number is simplified when it is generated. The equivalent mode is used to only calculate the number of changes between 0 and 1, thereby simplifying the complexity of the feature and obtaining a more robust and discriminative feature description. The LBP feature extraction effect is as follows: Figure 7 shown.

[0069] Following the above, we will define the image longitudinal position information features. Since we need to calculate the thickness of the submucosal layer in the pathological image of the gastric wall submucosal specimen, we need to segment the image of this part. Since the submucosal tissue has a relatively obvious position feature and is located at the bottom of the case tissue, we define the image longitudinal position information features, that is, generate a vertical coordinate feature map with the same size as the image, the effect is as follows Figure 8 shown.

[0070] Step S4, feature fusion: superimpose the denoised image, texture feature image and ordinate feature map obtained by preprocessing to form a superimposed image for segmentation. This method superimposes the images obtained from steps S1 to S3 above as input data for image segmentation, and the effect is as follows: Figure 9 As shown, the input is given to the following UNet neural network.

[0071] Step S5, Segmentation of Submucosal Pathological Tissue: Use the UNet neural network to train the superimposed image for image segmentation to extract the submucosal layer. UNet is a deep learning model architecture commonly used for image segmentation tasks. The model architecture of the UNet neural network includes:

[0072] Convolution module: The convolution module includes two convolution operations, and each convolution is followed by batch normalization and ReLU activation function processing.

[0073] Downsampling path: Convolutional modules and max pooling are used to reduce the feature map resolution and increase the number of channels. During downsampling, a series of convolutional modules and max pooling operations gradually reduce the feature map resolution while increasing the number of channels, thereby extracting high-level, abstract features of the image. Specifically, starting with the number of input channels, through multiple downsampling steps, the number of channels gradually increases, while the size of the feature map gradually decreases.

[0074] Bottleneck layer: Features are further processed and compressed through convolution modules to obtain the most representative high-dimensional feature representation.

[0075] Feature splicing: The UNet model splices the features from the downsampling process onto the upsampled features according to the feature hierarchy for subsequent feature upsampling. This process aims to fuse the coarser but semantically rich features obtained from upsampling with the finer but less semantically rich features preserved during downsampling, thereby comprehensively utilizing feature information from different levels to improve the model's segmentation performance. This allows the model to better capture both the global structure and local details of the image, resulting in more accurate segmentation results.

[0076] Upsampling path: During the upsampling process, the deconvolution operation is used to gradually restore the resolution of the feature map. At the same time, the upsampled features are fused with the corresponding downsampling layer features through feature splicing, and then the features are integrated and optimized through the convolution module.

[0077] Output layer: A convolutional layer is used to map the features to the number of output channels. This method is a two-class image segmentation method with an output channel of 1. The Sigmoid activation function is used to limit the output value between 0 and 1 to obtain the probability that each pixel belongs to the target category.

[0078] In this embodiment, the model architecture of the UNet neural network also includes a position attention mechanism. Since the submucosal layer has a significant position feature in the image, that is, it appears in the middle and lower part of the image, in order to enhance the model with position attention, it is used to improve the model's sensitivity to position features and improve the model's segmentation accuracy. Position attention mechanism: It is used to encode and weight the position information of the feature map before feature fusion or skip connection, so as to improve the model's sensitivity to position features and improve the model's segmentation accuracy. In the step of weighting the position information of the feature map: first, the input feature map is linearly transformed to obtain the query vector Q, key vector K and value vector V; then, the dot product of Q and K is calculated by matrix multiplication, and then scaling and Softmax operation are performed to obtain the attention weight matrix A; finally, the attention weight matrix A is multiplied by the value vector V to obtain the weighted feature representation to enhance the model's recognition ability of the region of interest.

[0079] In the steps of training the image segmentation task of superimposed images using the UNet neural network: Model training: and using the stochastic gradient descent algorithm to optimize the model parameters to adapt to image segmentation under the working conditions of 5-channel image feature input.

[0080] As mentioned above, the segmented image is processed: False positive and false negative points in the segmentation results are corrected using connected region area filtering. Holes in false negative points are closed using a combination of hole filling and closing operations, and edge smoothing is used to smooth the edges of false negative points to ensure the accuracy of the segmentation results. False positive points typically appear as erroneous discrete points. To correct them, connected region area filtering is used to correct false positive points. A connected region is a region in the image consisting of adjacent pixels with the same attribute value. I(x,y) represents the segmentation result image, (x,y) represents the coordinates of the pixel points, and for each connected region, its area A can be calculated. The region with the largest area among all connected regions is denoted as A{max}. When correcting false positive points, if there is a connected region with an area Ai smaller than the area of the largest connected region A{max} and is marked as positive, the corresponding connected region is removed from the positive results.

[0081]

[0082] In this way, the false positive results are corrected, thereby improving the accuracy of the segmentation results.

[0083] False negative points exist in the segmentation results as holes and missing edges. Holes of false negative points in the segmented image are first closed using hole filling combined with closing operations, and then edge smoothing operations are used to address false negative errors at the edges of false negative points in the segmented image. The hole filling algorithm: 1. Determine the hole area: traverse each negative connected area in the segmented image. For each negative connected area, check the pixels in the 8 connected areas outside it. If all pixels are positive, mark the negative connected area as a hole area; 2. Hole filling: Modify all negative pixel values in each marked hole area to positive pixel values. When determining the hole area, the specific method for checking the 8-connected area is: for the edge pixel (x, y) of the current negative connected area, the pixel coordinates of its 8-connected area are (x - 1, y - 1), (x - 1, y), (x - 1, y + 1), (x, y - 1), (x, y + 1), (x + 1, y - 1), (x + 1, y), (x + 1, y + 1). The pixel values at these positions are checked in turn to see if all pixels that are not inside the connected area are positive.

[0084] Step S6, Multi-point Thickness Calculation: A skeletonization method is used to extract the submucosal skeleton. This skeleton serves as the submucosal extension direction at that location. A perpendicular line is used to calculate the submucosal thickness at that location, enabling multi-point thickness measurement of the submucosal layer during endoscopic submucosal dissection. Since the segmented image in step S5 represents the submucosal tissue in the pathological image, the multi-point submucosal thickness measurement method measures the thickness of this segmentation result. Since the submucosal layer in the pathological image is not strictly horizontal, accurate submucosal thickness and determination of the submucosal tissue extension direction are necessary.

[0085] Step S7, Quality Assessment: The image is vertically divided into equal parts by manually selecting line endpoints and performing equal divisions to assess surgical quality. The manually defined line endpoints are used as the four or five equal division points, and the image is divided into equal parts using lines perpendicular to the line at these points. If the manually selected points lie outside the segmentation result, the line is subtracted in that direction by up to 5% to remove tissue outside the segmentation result.

[0086] In summary, this application provides a multi-point thickness measurement method for the submucosal layer during endoscopic submucosal dissection. This method preprocesses the pathological images of surgically dissected tissue, then uses a variety of feature extraction methods for feature enhancement. Finally, a convolutional neural network is used to segment the submucosal layer. Closing operations and connected region filtering are used to remove erroneous results. Combined with a backbone orientation method, this method ultimately achieves multi-point thickness measurement of the submucosal layer of pathological tissue. Thus, this application uses pathological tissue images to measure submucosal thickness, which is helpful for evaluating the effectiveness of endoscopic submucosal dissection (ESD) and controlling surgical quality.

[0087] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0088] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of this application, ordinary technicians in this field can also make many forms without departing from the purpose of this application and the scope of protection of the claims, all of which are within the protection of this application.

Claims

1. A method for measuring the thickness of the submucosal layer at multiple points for endoscopic submucosal dissection, characterized in that: The following steps are involved: Image acquisition: Acquisition of pathological images of the cuts made by endoscopic submucosal dissection; Image preprocessing: improving the brightness and highlight contrast of the pathological image, and smoothing the pathological image through Gaussian filtering; Image feature extraction: using a local binary pattern (LBP) algorithm to extract texture features of the pathological image and generate a vertical coordinate feature map of the same size as the pathological image, wherein the vertical coordinate feature map is used to define the longitudinal position information features of the image; feature Fusion: Superimpose the denoised image, texture feature image and ordinate feature map obtained by preprocessing to form a superimposed image for segmentation; Submucosal pathological tissue segmentation: The superimposed images were trained using a UNet neural network for image segmentation to extract the submucosal layer. False positive and false negative points in the segmentation results were corrected using connected region area filtering. Holes in false negative points were closed using a combination of hole filling and closing operations, and edge smoothing was used to process the edges of false negative points to ensure the accuracy of the segmentation results. Multi-point thickness calculation: The skeleton of the submucosal layer is extracted using a skeletonization method. This skeleton is used as the extension direction of the submucosal layer at the corresponding position, and the thickness of the submucosal layer at that position is calculated using a vertical line. This allows for multi-point thickness measurement of the submucosal layer during endoscopic submucosal dissection. Quality assessment: The image was divided into equal parts vertically by manually selecting the endpoints of the line and performing an equal division operation to assess the surgical quality.

2. The method for measuring the multi-point thickness of the submucosal layer for endoscopic submucosal dissection according to claim 1, characterized in that: In the step of improving the brightness of the pathological image: The background brightness of the pathological image is adjusted, and the brightness of the V channel of the image is adjusted in the HSV color space, while the hue (H) and saturation (S) channels are retained to avoid image color distortion. The clipping histogram is defined as follows: , represents the original histogram, , clipping threshold Defined as 180.

3. The method for measuring the multi-point thickness of the submucosal layer for endoscopic submucosal dissection according to claim 1, characterized in that: In the step of improving the contrast of the bright part of the pathological image: Gamma transform is used to perform nonlinear transformation on the image pixel values to enhance the contrast of the bright part of the pathological image. The formula of gamma transform is: , A is the transformation coefficient, which is used to restore the transformed image to the original image bit space, γ is used as the exponent of the exponential calculation, and when γ is greater than 1, the image contrast in high brightness areas can be improved.

4. The method for measuring the multi-point thickness of the submucosal layer for endoscopic submucosal dissection according to claim 1, wherein: In the step of smoothing the pathological image by Gaussian filtering: The pathological image is processed by applying Gaussian filtering, using a standard deviation value to smooth the image and reduce the interference of noise and details, wherein the Gaussian filtering formula is: , G(x,y) is the image value after Gaussian filtering, x and y are the image pixel coordinates, and σ is the standard deviation of the Gaussian distribution.

5. The method for measuring the multi-point thickness of the submucosal layer for endoscopic submucosal dissection according to claim 1, characterized in that: In the step of extracting texture features of the pathological image: converting the color image of the pathological image into a grayscale image; The LBP algorithm is used to process the grayscale image. The LBP algorithm generates the texture feature map by comparing the grayscale values of each pixel with its neighboring pixels.

6. The method for measuring the multi-point thickness of the submucosal layer for endoscopic submucosal dissection according to claim 1, characterized in that: The model architecture of the UNet neural network includes: Convolution module: It includes two convolution operations, each of which is followed by batch normalization and ReLU activation function processing; Downsampling path: Reduce the feature map resolution and increase the number of channels through convolution modules and maximum pooling operations; Bottleneck layer: further processes and compresses features through convolution modules to obtain the most representative high-dimensional feature representation; Feature splicing: The UNet model splices the features in the downsampling process onto the upsampled features according to the feature hierarchy for subsequent feature upsampling. Upsampling path: During the upsampling process, deconvolution operations are used to gradually restore the resolution of the feature map. At the same time, the upsampled features are fused with the corresponding downsampled layer features through feature concatenation, and then the features are integrated and optimized through the convolution module. Output layer: A convolutional layer is used to map the features to the number of output channels. This method is a two-class image segmentation method with an output channel of 1. The Sigmoid activation function is used to limit the output value between 0 and 1 to obtain the probability that each pixel belongs to the target category.

7. The method for measuring the multi-point thickness of the submucosal layer for endoscopic submucosal dissection according to claim 6, characterized in that: The model architecture of the UNet neural network also includes: Position attention mechanism: used to encode and weight the position information of the feature map before feature fusion or jump connection, so as to improve the model's sensitivity to position features and improve the model's segmentation accuracy.

8. The method for measuring the multi-point thickness of the submucosal layer for endoscopic submucosal dissection according to claim 7, characterized in that: In the step of weighting the position information of the feature map: Perform a linear transformation on the input feature map to obtain the query vector Q, key vector K and value vector V; Calculate the dot product of Q and K through matrix multiplication, and then perform scaling and Softmax operations to obtain the attention weight matrix A; Multiply the attention weight matrix A with the value vector V to obtain the weighted feature representation to enhance the model's ability to recognize the region of interest.

9. The method for measuring the multi-point thickness of the submucosal layer for endoscopic submucosal dissection according to claim 1, wherein: In the step of using the UNet neural network to train the superimposed image for image segmentation tasks: Model training: The error loss is calculated using a combination of Dice parameters and mean square error, and the stochastic gradient descent algorithm is used to optimize the model parameters to adapt to image segmentation under the working conditions of 5-channel image feature input; Processing segmented images: For false positive points and false negative points in the segmentation results, they are corrected by using the connected area filtering method, the holes of false negative points are closed by using hole filling combined with closing operations, and the edges of false negative points are processed by edge smoothing operations to ensure the accuracy of the segmentation results.

10. The method for measuring the multi-point thickness of the submucosal layer for endoscopic submucosal dissection according to claim 9, characterized in that: In the step of closing the holes of false negative points in the segmented image using hole filling combined with closing operation: Determine the hole area: traverse each negative connected area in the segmented image, and for each negative connected area, check the pixels in the outer 8 connected areas. If all pixels are positive, mark the negative connected area as a hole area; Hole filling: Modify all negative pixel values in each marked hole area to positive pixel values.

Images