A method and system for recommending personalized treatment plans for extraperitoneal hernia

Abstract

The invention discloses a method and system for recommending personalized treatment plans for extra-abdominal hernia, and relates to the technical field of medical data processing, including detecting motion artifacts on medical images, and if there are motion artifacts, suppressing motion artifacts on the medical images; preprocessing the medical images through similarity and image-related data; re-identifying the extra-abdominal hernia area in the preprocessed medical images, and performing image segmentation on the extra-abdominal hernia area, generating third-related information of all extra-abdominal hernia structures based on first-related information and second-related information of some extra-abdominal hernia structures; determining the type and size of extra-abdominal hernia; determining the treatment plan according to basic data, basic symptoms, type and size of extra-abdominal hernia, and mapping the treatment plan on a visual interface to help doctors make reference. The quality of medical images is improved, a standard for recommending a unified treatment plan is formulated, doctors are assisted in making references, and the adaptability of the treatment plan is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of medical data processing, and in particular to a method and system for recommending a personalized treatment plan for extraperitoneal hernia. Background Art

[0002] Extraperitoneal hernia refers to the protrusion of some organs or tissues in the abdominal cavity through some congenital or acquired weak points, defects or holes in the abdominal wall to the surface of the body, forming a visible mass. This pathological phenomenon is commonly known as "hernia".

[0003] Doctors face a series of comprehensive challenges when formulating treatment plans for extraperitoneal hernias. These challenges stem from the diversity of individual differences among patients, including considerations of age, physical condition, underlying diseases, and other factors. At the same time, extraperitoneal hernias themselves are complex diseases, and their type, location, and size all have an important impact on the choice of treatment plans.

[0004] In the existing technology, different doctors may obtain different treatment plans when analyzing medical images with the naked eye and combining them with the patient's condition, resulting in low adaptability and accuracy of the treatment plans. There is an urgent need for a unified treatment plan standard to assist doctors in treatment and help doctors provide reliable reference. Summary of the invention

[0005] The purpose of the present invention is to solve the problem that the standards of treatment plans in the prior art are not unified, resulting in low adaptability and accuracy of treatment plans, and to propose a method for recommending personalized treatment plans for extraperitoneal hernia.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A personalized treatment plan for extraperitoneal hernia is recommended, including:

[0008] Acquire physiological data and imaging data of patients with extra-abdominal hernia. The physiological data include intra-abdominal pressure-related data, basic data, basic symptoms and imaging-related data. Define a cause degree through the intra-abdominal pressure-related data and basic data. The imaging data include medical images and external photographs of the affected part. Perform motion artifact detection on the medical images to determine whether there is motion artifact in the medical images. If there is motion artifact, perform motion artifact suppression on the medical images.

[0009] Identify the extra-abdominal hernia area and the extra-abdominal hernia contents in the medical image, determine the surrounding area of ​​the extra-abdominal hernia area by the degree of inducement, extract the image features of the extra-abdominal hernia area and the surrounding area respectively, calculate the similarity between the extra-abdominal hernia area and the surrounding area according to the image features, and pre-process the medical image by the similarity and image-related data;

[0010] Re-identify the extra-abdominal hernia region in the pre-processed medical image, and perform image segmentation on the extra-abdominal hernia region to obtain multiple extra-abdominal hernia partial regions, determine first relevant information of each extra-abdominal hernia structure according to the multiple extra-abdominal hernia partial regions, extract second relevant information of the partial extra-abdominal hernia structure through external photography of the affected part, and generate third relevant information of all extra-abdominal hernia structures based on the first relevant information and the second relevant information of the partial extra-abdominal hernia structure;

[0011] The structure of extra-abdominal hernia includes hernia ring and hernia sac. By inputting the third-party related information of hernia ring and hernia sac and the hernia contents of extra-abdominal hernia into the preset classification model, the type of extra-abdominal hernia is output, and the size of extra-abdominal hernia is obtained by integrating the third-party related information of the structure of extra-abdominal hernia.

[0012] The treatment plan is determined based on basic data, basic symptoms, type of extraperitoneal hernia and size of extraperitoneal hernia, and the treatment plan is mapped on a visual interface to help doctors for reference.

[0013] In some embodiments of the present application, a predisposition degree is defined by intra-abdominal pressure related data and basic data, including:

[0014] The intra-abdominal pressure-related data include abdominal wall strength and intra-abdominal pressure. The basic data types that affect the abdominal wall strength and intra-abdominal pressure are screened out from the basic data, and the characteristic values ​​of the abdominal wall strength and the intra-abdominal pressure are counted. The characteristic value is the maximum abdominal wall strength reduction or the maximum intra-abdominal pressure increase.

[0015] The degree of inducement is defined according to the types of basic data after screening, the characteristic values ​​of abdominal wall strength and the characteristic values ​​of intra-abdominal pressure;

[0016] ;in, For the degree of inducement, , are the influence weights of abdominal wall strength and intra-abdominal pressure, , They are the maximum decrease in abdominal wall strength and the maximum increase in intra-abdominal pressure, is the number of basic data types after screening, For the The influence weight of basic data, For the The size of the basic data, is a preset constant.

[0017] In some embodiments of the present application, if there is a motion artifact phenomenon, the motion artifact suppression is performed on the medical image, including:

[0018] An artifact region is identified in the medical image by using a preset artifact detection algorithm, and recorded as a first artifact region;

[0019] Determine the timestamp of the medical image, and obtain basic data in the same time period according to the timestamp of the medical image, set a normal range of the basic data, and mark the basic data that exceeds the normal range in the time period to obtain an abnormal range part, and find the area in the corresponding medical image according to the timestamp of the abnormal range part, which is recorded as the second artifact area;

[0020] The union of the first artifact region and the second artifact region is taken as the artifact region, and motion artifacts of the medical image are suppressed through the abnormal range of the basic data.

[0021] In some embodiments of the present application, the surrounding area of ​​the extraperitoneal hernia area is determined by the degree of inducement, including:

[0022] Calculate the maximum transverse diameter, maximum longitudinal diameter and area of ​​the extraperitoneal hernia region, and mark the two end points of the maximum transverse diameter and maximum longitudinal diameter at the edge of the extraperitoneal hernia region;

[0023] The lateral extension length and the longitudinal extension length are obtained by area mapping of the extraperitoneal hernia region, and the lateral extension length and the longitudinal extension length are adjusted according to the degree of inducement. The maximum transverse diameter and the maximum longitudinal diameter are expanded based on the two endpoints of each of the maximum transverse diameter and the maximum longitudinal diameter after adjustment, so as to determine the surrounding area of ​​the extraperitoneal hernia region according to the expanded area.

[0024] In some embodiments of the present application, the similarity between the extraperitoneal hernia region and its surrounding area is calculated based on image features, and the medical image is preprocessed by the similarity and image-related data, including:

[0025] Taking the center point of the extra-abdominal hernia region as the starting point, the extra-abdominal hernia region and its surrounding area are divided into several comparison areas according to the area of ​​the extra-abdominal hernia region, and each comparison area includes a partial extra-abdominal hernia region and a partial surrounding area;

[0026] The similarity of each image feature between part of the extraperitoneal hernia region and part of its surrounding region in each comparison region is calculated, and the similarity of all image features is integrated to obtain the overall similarity of each comparison region;

[0027] The similarity between the complete extraperitoneal hernia area and its surrounding area was determined by the overall similarity of each comparison area;

[0028] ;in, is the similarity between the complete extraperitoneal hernia area and its surrounding area, is the number of comparison areas, For the The overall similarity of the comparison area, For the The area of ​​the partial extraperitoneal hernia in the comparison area, For the The area of ​​the surrounding area within the comparison area, is a preset constant;

[0029] The preprocessing level is determined by combining the similarity and image-related data, and the medical images are preprocessed according to the preprocessing level.

[0030] In some embodiments of the present application, the third relevant information of all extra-abdominal hernia structures is generated based on the first relevant information and the second relevant information of some extra-abdominal hernia structures, including:

[0031] The first relevant information and the second relevant information both include size and position, and the size and position are mapped in the same space and recorded as the first size and the second size, the first position and the second position, respectively;

[0032] Third information about the extra-abdominal hernia structure is generated by comparing the first size and the second size, the first position and the second position.

[0033] In some embodiments of the present application, third related information about the structure of an extra-abdominal hernia is generated by comparing the first size and the second size, the first position and the second position, including:

[0034] If the difference between the first size and the second size exceeds the size threshold, a size correction factor is determined by the difference between the difference and the size threshold, and the sizes of the first relevant information of all extraperitoneal hernia structures are corrected by the size correction factor to generate the size of the third relevant information;

[0035] If the distance between the first position and the second position exceeds the distance threshold, a position correction factor is determined by the difference between the distance and the distance threshold, and the positions of the first relevant information of all extra-abdominal hernia structures are corrected by the position correction factor to generate the position of the third relevant information;

[0036] The information except the size and the position in the first relevant information of all extraperitoneal hernia structures is used as the third relevant information.

[0037] In some embodiments of the present application, third related information about the structure of an extra-abdominal hernia is generated by comparing the first size and the second size, the first position and the second position, including:

[0038] If the difference between the first size and the second size does not exceed the size threshold, or the distance between the first position and the second position does not exceed the distance threshold, the first relevant information of all extraperitoneal hernia structures is directly used as the third relevant information.

[0039] Correspondingly, the present application also provides a personalized treatment plan recommendation system for extra-abdominal hernia, including:

[0040] A collection module is used to obtain physiological data and imaging data of patients with extra-abdominal hernia. The physiological data include intra-abdominal pressure-related data, basic data, basic symptoms and imaging-related data. A predisposing degree is defined by the intra-abdominal pressure-related data and basic data. The imaging data include medical images and external photographs of the affected part. Motion artifact detection is performed on the medical images to determine whether there is a motion artifact phenomenon in the medical images. If there is a motion artifact phenomenon, the motion artifact is suppressed on the medical images.

[0041] A preprocessing module is used to identify the extra-abdominal hernia area and the extra-abdominal hernia contents in the medical image, and determine the surrounding area of ​​the extra-abdominal hernia area according to the degree of inducement, respectively extract image features in the extra-abdominal hernia area and the surrounding area, calculate the similarity between the extra-abdominal hernia area and the surrounding area according to the image features, and preprocess the medical image according to the similarity and image-related data;

[0042] A generation module is used to re-identify the extra-abdominal hernia region in the pre-processed medical image, and perform image segmentation on the extra-abdominal hernia region to obtain multiple extra-abdominal hernia partial regions, determine first relevant information of each extra-abdominal hernia structure according to the multiple extra-abdominal hernia partial regions, extract second relevant information of the partial extra-abdominal hernia structure through external photography of the affected part, and generate third relevant information of all extra-abdominal hernia structures based on the first relevant information and the second relevant information of the partial extra-abdominal hernia structure;

[0043] A determination module is used for the structure of extra-abdominal hernia, including hernia ring and hernia sac. By inputting the third-party related information of the hernia ring and hernia sac and the hernia contents of extra-abdominal hernia into a preset classification model, the type of extra-abdominal hernia is output, and the size of the extra-abdominal hernia is obtained by integrating the third-party related information of the structure of extra-abdominal hernia;

[0044] The recommendation module is used to determine the treatment plan based on basic data, basic symptoms, type of extraperitoneal hernia and size of extraperitoneal hernia, and map the treatment plan on a visual interface to help doctors for reference.

[0045] By adopting the above technical scheme, the physiological data and imaging data of the patient with extra-abdominal hernia are obtained, the physiological data including intra-abdominal pressure related data, basic data, basic symptoms and imaging related data, a predisposing degree is defined by the intra-abdominal pressure related data and basic data, the imaging data including medical images and external photographs of the affected part, motion artifact detection is performed on the medical images to determine whether there is a motion artifact phenomenon in the medical images, and if there is a motion artifact phenomenon, the medical images are subjected to motion artifact suppression; the extra-abdominal hernia area and the extra-abdominal hernia contents are identified in the medical images, and the surrounding area of ​​the extra-abdominal hernia area is determined by the predisposing degree, the image features of the extra-abdominal hernia area and the surrounding area are respectively extracted, the similarity between the extra-abdominal hernia area and the surrounding area is calculated according to the image features, and the medical images are pre-processed by the similarity and the imaging related data; Re-identify the extra-abdominal hernia area in the pre-processed medical image, and perform image segmentation on the extra-abdominal hernia area to obtain multiple extra-abdominal hernia partial areas, determine the first relevant information of each extra-abdominal hernia structure according to the multiple extra-abdominal hernia partial areas, extract the second relevant information of the partial extra-abdominal hernia structure through external photography of the affected part, and generate the third relevant information of all extra-abdominal hernia structures based on the first relevant information and the second relevant information of the partial extra-abdominal hernia structure; the extra-abdominal hernia structure includes a hernia ring and a hernia sac, and the third relevant information of the hernia ring and the hernia sac and the extra-abdominal hernia contents are input into a preset classification model, and the type of extra-abdominal hernia is output, and the size of the extra-abdominal hernia is obtained according to the third relevant information of the extra-abdominal hernia structure; determine the treatment plan according to basic data, basic symptoms, type of extra-abdominal hernia and size of extra-abdominal hernia, and map the treatment plan on a visual interface to help doctors for reference. The present application improves the quality of medical images by suppressing motion artifacts in medical images and pre-processing medical images through similarity and image-related data, thereby improving the accuracy of image processing. The third relevant information of all extra-abdominal hernia structures is generated based on the first relevant information and the second relevant information of the partial extra-abdominal hernia structure, and the first relevant information is adjusted to ensure the accuracy of the relevant information of the extra-abdominal hernia structure. The treatment plan is determined based on basic data, basic symptoms, type of extraperitoneal hernia and size of extraperitoneal hernia. A unified standard for recommending treatment plans is formulated to assist doctors in their reference, ensuring the adaptability of the treatment plan. It also saves analysis time and improves the doctor's work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 A schematic diagram of a process for recommending a personalized treatment plan for extra-abdominal hernia proposed by the present invention;

[0047] Figure 2 This is a schematic diagram of the structure of a personalized treatment plan recommendation system for extraperitoneal hernia proposed by the present invention. DETAILED DESCRIPTION

[0048] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0049] Reference Figure 1 , a method for recommending a personalized treatment plan for extraperitoneal hernia, comprising the following steps:

[0050] Step S101, obtaining physiological data and imaging data of a patient with extra-abdominal hernia, wherein the physiological data include intra-abdominal pressure-related data, basic data, basic symptoms and imaging-related data, and defining a predisposing degree through the intra-abdominal pressure-related data and basic data; the imaging data include medical images and external photographs of the affected part, and performing motion artifact detection on the medical images to determine whether there is motion artifact in the medical images; if there is motion artifact, performing motion artifact suppression on the medical images.

[0051] In this embodiment, intra-abdominal pressure related data (abdominal wall strength and intra-abdominal pressure), basic data (age, height, weight, etc.), basic symptoms (pain, nausea, etc.) and image related data (physiological data that affect image analysis, such as body fat). The direct causes of the formation of extra-abdominal hernia mainly include two factors: reduced abdominal wall strength and increased intra-abdominal pressure. Therefore, these two factors are considered to define a predisposing degree, which describes the degree of continuous influence on extra-abdominal hernia. Combined with the patient's physiological data, motion artifacts are suppressed in medical images to improve the quality of medical images.

[0052] In some embodiments of the present application, a predisposition degree is defined by intra-abdominal pressure related data and basic data, including:

[0053] The intra-abdominal pressure-related data include abdominal wall strength and intra-abdominal pressure. The basic data types that affect the abdominal wall strength and intra-abdominal pressure are screened out from the basic data, and the characteristic values ​​of the abdominal wall strength and the intra-abdominal pressure are counted. The characteristic value is the maximum abdominal wall strength reduction or the maximum intra-abdominal pressure increase.

[0054] The degree of inducement is defined according to the types of basic data after screening, the characteristic values ​​of abdominal wall strength and the characteristic values ​​of intra-abdominal pressure;

[0055] ;in, For the degree of inducement, , are the influence weights of abdominal wall strength and intra-abdominal pressure, , They are the maximum decrease in abdominal wall strength and the maximum increase in intra-abdominal pressure, is the number of basic data types after screening, For the The influence weight of basic data, For the The size of the basic data, is a preset constant.

[0056] In this embodiment, the characteristic values ​​of abdominal wall strength and intra-abdominal pressure are counted, and the data used for the statistics are data within the current period of time. The abdominal wall strength can be quantified by evaluating the integrity of the abdominal wall tissue, the strength of the muscles and fascia, and the intra-abdominal pressure can be evaluated by monitoring the changes in the intra-abdominal pressure. There are some basic data that may affect the abdominal wall strength and intra-abdominal pressure, such as age, weight, etc. The effects of the abdominal wall strength and intra-abdominal pressure are corrected by these basic data.

[0057] In some embodiments of the present application, if there is a motion artifact phenomenon, the motion artifact suppression is performed on the medical image, including:

[0058] An artifact region is identified in the medical image by using a preset artifact detection algorithm, and recorded as a first artifact region;

[0059] Determine the timestamp of the medical image, and obtain basic data in the same time period according to the timestamp of the medical image, set a normal range of the basic data, and mark the basic data that exceeds the normal range in the time period to obtain an abnormal range part, and find the area in the corresponding medical image according to the timestamp of the abnormal range part, which is recorded as the second artifact area;

[0060] The union of the first artifact region and the second artifact region is taken as the artifact region, and motion artifacts of the medical image are suppressed through the abnormal range of the basic data.

[0061] In this embodiment, ensure that basic data (such as electrocardiogram, respiratory signal, etc.) synchronized with image acquisition is available. Analyze the basic data to identify abnormal patterns or events related to motion artifacts, such as abnormal stages in the heartbeat and respiratory cycles, and determine the corresponding positions of these physiological events on the image acquisition timeline. Work by analyzing the image and basic data in the frequency domain. Specific frequency components in the basic data (such as electrocardiogram) may correspond to the frequency components of motion artifacts in the image. Design filters through the abnormal range part to suppress these frequency components related to physiological events, thereby reducing motion artifacts in the image.

[0062] Step S102, identifying the extra-abdominal hernia area and the extra-abdominal hernia contents in the medical image, and determining the surrounding area of ​​the extra-abdominal hernia area by the degree of inducement, respectively extracting image features in the extra-abdominal hernia area and the surrounding area, calculating the similarity between the extra-abdominal hernia area and the surrounding area according to the image features, and pre-processing the medical image by the similarity and image-related data.

[0063] In this embodiment, a preliminary blur detection is performed on the medical image, and the target detection algorithm (such as Faster R-CNN, YOLO, etc.) is used for identification. Because in some cases, the extraperitoneal hernia may be very hidden or the contents of the hernia sac are very similar to the surrounding tissues, making it difficult for the algorithm to accurately distinguish them, the medical image is preprocessed here according to the similarity and image-related data, such as increasing the contrast to the target value to facilitate distinction. The surrounding area is defined by the sustainable impact of the cause.

[0064] In some embodiments of the present application, the surrounding area of ​​the extraperitoneal hernia area is determined by the degree of inducement, including:

[0065] Calculate the maximum transverse diameter, maximum longitudinal diameter and area of ​​the extraperitoneal hernia region, and mark the two end points of the maximum transverse diameter and maximum longitudinal diameter at the edge of the extraperitoneal hernia region;

[0066] The lateral extension length and the longitudinal extension length are obtained by area mapping of the extraperitoneal hernia region, and the lateral extension length and the longitudinal extension length are adjusted according to the degree of inducement. The maximum transverse diameter and the maximum longitudinal diameter are expanded based on the two endpoints of each of the maximum transverse diameter and the maximum longitudinal diameter after adjustment, so as to determine the surrounding area of ​​the extraperitoneal hernia region according to the expanded area.

[0067] In this embodiment, different areas correspond to different lateral expansion lengths and longitudinal expansion lengths. Different inducement degrees correspond to different lateral expansion coefficients and longitudinal expansion coefficients. The lateral expansion length after expansion is obtained by multiplying the lateral expansion length by the expansion coefficient, and the same is true for the longitudinal expansion.

[0068] In some embodiments of the present application, the similarity between the extraperitoneal hernia region and its surrounding area is calculated based on image features, and the medical image is preprocessed by the similarity and image-related data, including:

[0069] Taking the center point of the extra-abdominal hernia region as the starting point, the extra-abdominal hernia region and its surrounding area are divided into several comparison areas according to the area of ​​the extra-abdominal hernia region, and each comparison area includes a partial extra-abdominal hernia region and a partial surrounding area;

[0070] The similarity of each image feature between part of the extraperitoneal hernia region and part of its surrounding region in each comparison region is calculated, and the similarity of all image features is integrated to obtain the overall similarity of each comparison region;

[0071] The similarity between the complete extraperitoneal hernia area and its surrounding area was determined by the overall similarity of each comparison area;

[0072] ;in, is the similarity between the complete extraperitoneal hernia area and its surrounding area, is the number of comparison areas, For the The overall similarity of the comparison area, For the The area of ​​the partial extraperitoneal hernia in the comparison area, For the The area of ​​the surrounding area within the comparison area, is a preset constant;

[0073] The preprocessing level is determined by combining the similarity and image-related data, and the medical images are preprocessed according to the preprocessing level.

[0074] In this embodiment, the similarity of all image features can be integrated by weighted summation or weighted average. The overall similarity is corrected according to the area difference between the part of the extraperitoneal hernia area and the part of the surrounding area in the comparison area.

[0075] In this embodiment, the preprocessing level is determined in combination with the similarity and image-related data. Some physiological data will affect image analysis, such as fat, etc. The preprocessing level is determined based on the similarity and these image-related data (which can be determined by a machine learning model). Different preprocessing levels correspond to different contrast target values, denoising target values, sharpening target values, etc. Medical images are preprocessed using these target values.

[0076] Step S103, re-identify the extra-abdominal hernia area in the pre-processed medical image, and perform image segmentation on the extra-abdominal hernia area to obtain multiple extra-abdominal hernia partial areas, determine the first relevant information of each extra-abdominal hernia structure according to the multiple extra-abdominal hernia partial areas, extract the second relevant information of the partial extra-abdominal hernia structure through external photography of the affected area, and generate the third relevant information of all extra-abdominal hernia structures based on the first relevant information and the second relevant information of the partial extra-abdominal hernia structure.

[0077] In this embodiment, the extra-abdominal hernia region is identified in the high-quality medical image after pretreatment, and image segmentation is performed to identify each extra-abdominal hernia partial region. The typical extra-abdominal hernia structure includes a hernia ring, a hernia sac, and a hernia outer cover. The hernia ring is the place where the hernia sac protrudes from the abdominal cavity, that is, the weak area or defective part of the abdominal wall. Various types of hernias are named according to the location of the hernia ring, such as inguinal hernia, umbilical hernia, femoral hernia, etc. The hernia sac is a sac structure protruding from the peritoneal wall through the hernia ring, which can be divided into three parts: the sac neck, the sac body, and the sac bottom. The sac neck is the narrow part where the hernia sac is connected to the abdominal cavity, which is thickened by friction. The hernia contents refer to the organs and tissues that protrude from the abdominal cavity into the hernia sac. The most common ones are the small intestine and the greater omentum. Others such as the cecum, appendix, sigmoid colon, etc. may also become hernia contents. The hernia outer cover refers to the various layers of tissues of the abdominal wall outside the hernia sac, which are usually composed of fascia, muscle, subcutaneous tissue and skin. The first relevant information includes size (which can be area or volume), position, shape, etc.

[0078] In this embodiment, the second relevant information of part of the extra-abdominal hernia structure is extracted by external photographic images of the affected part, which are not medical images but photos of the affected part taken by a photographing device from outside the patient's body. The second relevant information of part of the extra-abdominal hernia structure (hernial sac, hernial external cover) is extracted by image processing technology, and the second relevant information only includes size and position.

[0079] In this embodiment, the third relevant information of all extra-abdominal hernia structures is generated based on the first relevant information and the second relevant information of some extra-abdominal hernia structures, and the position and size of other extra-abdominal hernia structures are corrected by the position and size deviations on two types of images of some extra-abdominal hernia structures, so as to obtain the third relevant information of all extra-abdominal hernia structures.

[0080] In some embodiments of the present application, the third relevant information of all extra-abdominal hernia structures is generated based on the first relevant information and the second relevant information of some extra-abdominal hernia structures, including:

[0081] The first relevant information and the second relevant information both include size and position, and the size and position are mapped in the same space and recorded as the first size and the second size, the first position and the second position, respectively;

[0082] Third information about the extra-abdominal hernia structure is generated by comparing the first size and the second size, the first position and the second position.

[0083] In this embodiment, the size and position are mapped in the same space, and the size and position are determined in two-dimensional coordinates or three-dimensional coordinates.

[0084] In some embodiments of the present application, third related information about the structure of an extra-abdominal hernia is generated by comparing the first size and the second size, the first position and the second position, including:

[0085] If the difference between the first size and the second size exceeds the size threshold, a size correction factor is determined by the difference between the difference and the size threshold, and the sizes of the first relevant information of all extraperitoneal hernia structures are corrected by the size correction factor to generate the size of the third relevant information;

[0086] If the distance between the first position and the second position exceeds the distance threshold, a position correction factor is determined by the difference between the distance and the distance threshold, and the positions of the first relevant information of all extra-abdominal hernia structures are corrected by the position correction factor to generate the position of the third relevant information;

[0087] The information except the size and the position in the first relevant information of all extraperitoneal hernia structures is used as the third relevant information.

[0088] In this embodiment, if the deviation of size or position exceeds a reasonable range, correction is performed. The size correction factor is determined by the difference between the difference value and the size threshold value. Different difference values ​​correspond to different size correction factors. The size of the third related information is obtained by multiplying the size correction factor by the first size, and the same is true for the position correction factor.

[0089] In some embodiments of the present application, third related information about the structure of an extra-abdominal hernia is generated by comparing the first size and the second size, the first position and the second position, including:

[0090] If the difference between the first size and the second size does not exceed the size threshold, or the distance between the first position and the second position does not exceed the distance threshold, the first relevant information of all extraperitoneal hernia structures is directly used as the third relevant information.

[0091] In this embodiment, if the deviation in size or position does not exceed a reasonable range, the first relevant information is used as the third relevant information, and the above correction refers to correcting the position and size of the first relevant information based on the deviation in position and size between the second relevant information and the first relevant information.

[0092] Step S104, the extra-abdominal hernia structure includes a hernia ring and a hernia sac. By inputting the third-related information of the hernia ring and the hernia sac and the hernia contents of the extra-abdominal hernia into a preset classification model, the type of extra-abdominal hernia is output, and the size of the extra-abdominal hernia is obtained by integrating the third-related information of the extra-abdominal hernia structure.

[0093] In this embodiment, the type of extra-abdominal hernia mainly depends on the position of the hernia ring and the hernia sac and the contents of the extra-abdominal hernia, so as to determine the two most likely types of extra-abdominal hernia, which can be divided into multiple types such as inguinal hernia, umbilical hernia, and femoral hernia. The size of the extra-abdominal hernia is obtained by integrating the third-party related information of the extra-abdominal hernia structure, and three-dimensional modeling is performed based on the third-party related information of these extra-abdominal hernia structures, so as to determine the volume of each extra-abdominal hernia structure and obtain the size of the extra-abdominal hernia.

[0094] Step S105, determining a treatment plan based on the basic data, basic symptoms, type of extraperitoneal hernia and size of extraperitoneal hernia, and mapping the treatment plan on a visual interface to help doctors for reference.

[0095] In this embodiment, the two most likely types of extra-abdominal hernia are used to determine the treatment methods corresponding to each type. Some types of extra-abdominal hernia require surgical treatment, while some types of extra-abdominal hernia require conservative treatment. Basic data, basic symptoms, types of extra-abdominal hernia, and sizes of extra-abdominal hernia are input into a preset treatment plan recommendation model to determine various parameters involved in the treatment plan, and these contents are mapped on a visual interface for reference by doctors.

[0096] By adopting the above technical scheme, the physiological data and imaging data of the patient with extra-abdominal hernia are obtained, the physiological data including intra-abdominal pressure related data, basic data, basic symptoms and imaging related data, a predisposing degree is defined by the intra-abdominal pressure related data and basic data, the imaging data including medical images and external photographs of the affected part, motion artifact detection is performed on the medical images to determine whether there is a motion artifact phenomenon in the medical images, and if there is a motion artifact phenomenon, the medical images are subjected to motion artifact suppression; the extra-abdominal hernia area and the extra-abdominal hernia contents are identified in the medical images, and the surrounding area of ​​the extra-abdominal hernia area is determined by the predisposing degree, the image features of the extra-abdominal hernia area and the surrounding area are respectively extracted, the similarity between the extra-abdominal hernia area and the surrounding area is calculated according to the image features, and the medical images are pre-processed by the similarity and the imaging related data; Re-identify the extra-abdominal hernia area in the pre-processed medical image, and perform image segmentation on the extra-abdominal hernia area to obtain multiple extra-abdominal hernia partial areas, determine the first relevant information of each extra-abdominal hernia structure according to the multiple extra-abdominal hernia partial areas, extract the second relevant information of the partial extra-abdominal hernia structure through external photography of the affected part, and generate the third relevant information of all extra-abdominal hernia structures based on the first relevant information and the second relevant information of the partial extra-abdominal hernia structure; the extra-abdominal hernia structure includes a hernia ring and a hernia sac, and the third relevant information of the hernia ring and the hernia sac and the extra-abdominal hernia contents are input into a preset classification model, and the type of extra-abdominal hernia is output, and the size of the extra-abdominal hernia is obtained according to the third relevant information of the extra-abdominal hernia structure; determine the treatment plan according to basic data, basic symptoms, type of extra-abdominal hernia and size of extra-abdominal hernia, and map the treatment plan on a visual interface to help doctors for reference. The present application improves the quality of medical images by suppressing motion artifacts in medical images and pre-processing medical images through similarity and image-related data, thereby improving the accuracy of image processing. The third relevant information of all extra-abdominal hernia structures is generated based on the first relevant information and the second relevant information of the partial extra-abdominal hernia structure, and the first relevant information is adjusted to ensure the accuracy of the relevant information of the extra-abdominal hernia structure. The treatment plan is determined based on basic data, basic symptoms, type of extraperitoneal hernia and size of extraperitoneal hernia. A unified standard for recommending treatment plans is formulated to assist doctors in their reference, ensuring the adaptability of the treatment plan. It also saves analysis time and improves the doctor's work efficiency.

[0097] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present invention can be implemented by hardware, or by software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present invention.

[0098] Correspondingly, the present application also provides a personalized treatment plan recommendation system for extra-abdominal hernia, such as Figure 2 As shown, including:

[0099] A collection module is used to obtain physiological data and imaging data of patients with extra-abdominal hernia. The physiological data include intra-abdominal pressure-related data, basic data, basic symptoms and imaging-related data. A predisposing degree is defined by the intra-abdominal pressure-related data and basic data. The imaging data include medical images and external photographs of the affected part. Motion artifact detection is performed on the medical images to determine whether there is a motion artifact phenomenon in the medical images. If there is a motion artifact phenomenon, the motion artifact is suppressed on the medical images.

[0100] A preprocessing module is used to identify the extra-abdominal hernia area and the extra-abdominal hernia contents in the medical image, and determine the surrounding area of ​​the extra-abdominal hernia area according to the degree of inducement, respectively extract image features in the extra-abdominal hernia area and the surrounding area, calculate the similarity between the extra-abdominal hernia area and the surrounding area according to the image features, and preprocess the medical image according to the similarity and image-related data;

[0101] A generation module is used to re-identify the extra-abdominal hernia region in the pre-processed medical image, and perform image segmentation on the extra-abdominal hernia region to obtain multiple extra-abdominal hernia partial regions, determine first relevant information of each extra-abdominal hernia structure according to the multiple extra-abdominal hernia partial regions, extract second relevant information of the partial extra-abdominal hernia structure through external photography of the affected part, and generate third relevant information of all extra-abdominal hernia structures based on the first relevant information and the second relevant information of the partial extra-abdominal hernia structure;

[0102] A determination module is used for the structure of extra-abdominal hernia, including hernia ring and hernia sac. By inputting the third-party related information of the hernia ring and hernia sac and the hernia contents of extra-abdominal hernia into a preset classification model, the type of extra-abdominal hernia is output, and the size of the extra-abdominal hernia is obtained by integrating the third-party related information of the structure of extra-abdominal hernia;

[0103] The recommendation module is used to determine the treatment plan based on basic data, basic symptoms, type of extraperitoneal hernia and size of extraperitoneal hernia, and map the treatment plan on a visual interface to help doctors for reference.

[0104] Those skilled in the art will appreciate that the modules in the system in the implementation scenario can be distributed in the system of the implementation scenario according to the implementation scenario description, or can be changed accordingly and located in one or more systems different from the implementation scenario. The modules in the above implementation scenario can be combined into one module, or can be further split into multiple submodules.

[0105] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for recommending personalized treatment plans for extraperitoneal hernia, characterized in that: include: Acquire physiological data and imaging data of patients with extra-abdominal hernia. The physiological data include intra-abdominal pressure-related data, basic data, basic symptoms and imaging-related data. Define a cause degree through the intra-abdominal pressure-related data and basic data. The imaging data include medical images and external photographs of the affected part. Perform motion artifact detection on the medical images to determine whether there is motion artifact in the medical images. If there is motion artifact, perform motion artifact suppression on the medical images. Identify the extra-abdominal hernia area and the extra-abdominal hernia contents in the medical image, determine the surrounding area of ​​the extra-abdominal hernia area by the degree of inducement, extract the image features of the extra-abdominal hernia area and the surrounding area respectively, calculate the similarity between the extra-abdominal hernia area and the surrounding area according to the image features, and pre-process the medical image by the similarity and image-related data; Re-identify the extra-abdominal hernia region in the pre-processed medical image, and perform image segmentation on the extra-abdominal hernia region to obtain multiple extra-abdominal hernia partial regions, determine first relevant information of each extra-abdominal hernia structure according to the multiple extra-abdominal hernia partial regions, extract second relevant information of the partial extra-abdominal hernia structure through external photography of the affected part, and generate third relevant information of all extra-abdominal hernia structures based on the first relevant information and the second relevant information of the partial extra-abdominal hernia structure; The structure of extra-abdominal hernia includes hernia ring and hernia sac. By inputting the third-party related information of hernia ring and hernia sac and the hernia contents of extra-abdominal hernia into the preset classification model, the type of extra-abdominal hernia is output, and the size of extra-abdominal hernia is obtained by integrating the third-party related information of the structure of extra-abdominal hernia. Determine the treatment plan based on basic data, basic symptoms, type of extraperitoneal hernia and size of extraperitoneal hernia, and map the treatment plan on a visual interface to help doctors with reference; in, A predisposing degree is defined by intra-abdominal pressure related data and basic data, including: The intra-abdominal pressure-related data include abdominal wall strength and intra-abdominal pressure. The basic data types that affect the abdominal wall strength and intra-abdominal pressure are screened out from the basic data, and the characteristic values ​​of the abdominal wall strength and the intra-abdominal pressure are counted. The characteristic value is the maximum abdominal wall strength reduction or the maximum intra-abdominal pressure increase. The degree of inducement is defined according to the types of basic data after screening, the characteristic values ​​of abdominal wall strength and the characteristic values ​​of intra-abdominal pressure; ;in, For the degree of inducement, , are the influence weights of abdominal wall strength and intra-abdominal pressure, , They are the maximum decrease in abdominal wall strength and the maximum increase in intra-abdominal pressure, is the number of basic data types after screening, For the The influence weight of basic data, For the The size of the basic data, is a preset constant; If motion artifacts exist, the medical images are subjected to motion artifact suppression, including: An artifact region is identified in the medical image by using a preset artifact detection algorithm, and recorded as a first artifact region; Determine the timestamp of the medical image, and obtain basic data in the same time period according to the timestamp of the medical image, set a normal range of the basic data, and mark the basic data that exceeds the normal range in the time period to obtain an abnormal range part, and find the area in the corresponding medical image according to the timestamp of the abnormal range part, which is recorded as the second artifact area; The union of the first artifact region and the second artifact region is used as the artifact region, and motion artifacts are suppressed on the medical image through the abnormal range of the basic data; And the surrounding area of ​​the extraperitoneal hernia area is determined by the degree of predisposition, including: Calculate the maximum transverse diameter, maximum longitudinal diameter and area of ​​the extraperitoneal hernia region, and mark the two end points of the maximum transverse diameter and maximum longitudinal diameter at the edge of the extraperitoneal hernia region; The transverse extension length and the longitudinal extension length are obtained by area mapping of the extraperitoneal hernia region, and the transverse extension length and the longitudinal extension length are adjusted according to the inducement degree. The maximum transverse diameter and the maximum longitudinal diameter are expanded based on the two end points of each of the maximum transverse diameter and the maximum longitudinal diameter according to the adjusted transverse extension length and the longitudinal extension length, so as to determine the surrounding area of ​​the extraperitoneal hernia region according to the expanded area; The similarity between the extraperitoneal hernia area and its surrounding area is calculated based on the image features, and the medical image is preprocessed through the similarity and image-related data, including: Taking the center point of the extra-abdominal hernia region as the starting point, the extra-abdominal hernia region and its surrounding area are divided into several comparison areas according to the area of ​​the extra-abdominal hernia region, and each comparison area includes a partial extra-abdominal hernia region and a partial surrounding area; The similarity of each image feature between part of the extraperitoneal hernia region and part of its surrounding region in each comparison region is calculated, and the similarity of all image features is integrated to obtain the overall similarity of each comparison region; The similarity between the complete extraperitoneal hernia area and its surrounding area was determined by the overall similarity of each comparison area; ;in, is the similarity between the complete extraperitoneal hernia area and its surrounding area, is the number of comparison areas, For the The overall similarity of the comparison area, For the The area of ​​the partial extraperitoneal hernia in the comparison area, For the The area of ​​the surrounding area within the comparison area, is a preset constant; Determine the preprocessing level by combining the similarity and image-related data, and preprocess the medical image by using the preprocessing level; The third related information of all extra-abdominal hernia structures is generated based on the first related information and the second related information of some extra-abdominal hernia structures, including: The first relevant information and the second relevant information both include size and position, and the size and position are mapped in the same space and recorded as the first size and the second size, the first position and the second position, respectively; Third information about the extra-abdominal hernia structure is generated by comparing the first size and the second size, the first position and the second position.

2. A method for recommending personalized treatment plans for extraperitoneal hernia according to claim 1, characterized in that: Generating third related information about the structure of the extraperitoneal hernia by comparing the first size and the second size, the first position and the second position, includes: If the difference between the first size and the second size exceeds the size threshold, a size correction factor is determined by the difference between the difference and the size threshold, and the sizes of the first relevant information of all extraperitoneal hernia structures are corrected by the size correction factor to generate the size of the third relevant information; If the distance between the first position and the second position exceeds the distance threshold, a position correction factor is determined by the difference between the distance and the distance threshold, and the positions of the first relevant information of all extra-abdominal hernia structures are corrected by the position correction factor to generate the position of the third relevant information; The information except the size and the position in the first relevant information of all extraperitoneal hernia structures is used as the third relevant information.

3. A method for recommending personalized treatment plans for extraperitoneal hernia according to claim 2, characterized in that: Generating third related information about the structure of the extraperitoneal hernia by comparing the first size and the second size, the first position and the second position, includes: If the difference between the first size and the second size does not exceed the size threshold, or the distance between the first position and the second position does not exceed the distance threshold, the first relevant information of all extraperitoneal hernia structures is directly used as the third relevant information.

4. A personalized treatment plan recommendation system for extra-abdominal hernia, used to implement the personalized treatment plan recommendation method for extra-abdominal hernia according to any one of claims 1 to 3, characterized in that: include: A collection module is used to obtain physiological data and imaging data of patients with extra-abdominal hernia. The physiological data include intra-abdominal pressure-related data, basic data, basic symptoms and imaging-related data. A predisposing degree is defined by the intra-abdominal pressure-related data and basic data. The imaging data include medical images and external photographs of the affected part. Motion artifact detection is performed on the medical images to determine whether there is a motion artifact phenomenon in the medical images. If there is a motion artifact phenomenon, the motion artifact is suppressed on the medical images. A preprocessing module is used to identify the extra-abdominal hernia area and the extra-abdominal hernia contents in the medical image, and determine the surrounding area of ​​the extra-abdominal hernia area according to the degree of inducement, respectively extract image features in the extra-abdominal hernia area and the surrounding area, calculate the similarity between the extra-abdominal hernia area and the surrounding area according to the image features, and preprocess the medical image according to the similarity and image-related data; A generation module is used to re-identify the extra-abdominal hernia region in the pre-processed medical image, and perform image segmentation on the extra-abdominal hernia region to obtain multiple extra-abdominal hernia partial regions, determine first relevant information of each extra-abdominal hernia structure according to the multiple extra-abdominal hernia partial regions, extract second relevant information of the partial extra-abdominal hernia structure through external photography of the affected part, and generate third relevant information of all extra-abdominal hernia structures based on the first relevant information and the second relevant information of the partial extra-abdominal hernia structure; A determination module is used for the structure of extra-abdominal hernia, including hernia ring and hernia sac. By inputting the third-party related information of the hernia ring and hernia sac and the hernia contents of extra-abdominal hernia into a preset classification model, the type of extra-abdominal hernia is output, and the size of the extra-abdominal hernia is obtained by integrating the third-party related information of the structure of extra-abdominal hernia; The recommendation module is used to determine the treatment plan based on basic data, basic symptoms, type of extraperitoneal hernia and size of extraperitoneal hernia, and map the treatment plan on a visual interface to help doctors for reference.

Images