Microsatellite stability state prediction system and method for rectal cancer patient

Abstract

The invention belongs to the technical field of medical treatment, and discloses a rectal cancer patient microsatellite stability state prediction system and method. Comprising a training data set generation module, a microsatellite stability state evaluation module, an image acquisition and segmentation module, a radiomics feature extraction and screening module, a stability state prediction model construction module and a microsatellite stability state prediction module, and the radiomics feature extraction and screening module is used for extracting radiomics features of each tumor area in the three-dimensional volume diagram, and screening the extracted radiomics features based on a normal distribution method and a minimum absolute contraction and selection operator to obtain key features for predicting the stability state of the microsatellite. The image omics provides a non-invasive alternative technology, the stability state of the microsatellite can be predicted through image data, and the pain and risk of a patient are reduced.

Description

Technical Field

[0001] The present invention relates to the field of medical technology, and in particular to a system and method for predicting the stability state of microsatellites in patients with rectal cancer. Background Art

[0002] Colorectal cancer is a highly heterogeneous and dynamic disease with multiple molecular pathways involved in its development, and each patient has differences in clinical manifestations, treatment response, and prognosis. Microsatellite stability status is maintained by mismatch repair genes. When the mismatch repair system is impaired, defective microsatellite sequences accumulate, leading to microsatellite instability and early-onset colorectal cancer. The microsatellite stability status of colorectal cancer is closely related to patient prognosis and chemotherapy response. Therefore, professional organizations for cancer research and treatment recommend that all colorectal cancer patients undergo microsatellite stability status testing.

[0003] Currently, the main methods for detecting microsatellite stability status are immunohistochemistry, polymerase chain reaction, and high-throughput sequencing. However, these methods can only be determined by preoperative biopsy or postoperative tissue specimens. Due to the spatial and temporal heterogeneity of tumors, the results of microsatellite stability status detection may vary depending on the location and time of biopsy. In addition, biopsy is an invasive procedure that may cause pain, infection or other complications to patients. Magnetic resonance imaging has the characteristics of non-invasive, high-quality soft tissue imaging, multi-parameter, and multi-directional imaging. It can clearly display the soft tissue structure of each layer of the rectal wall, better evaluate the clinical stage and heterogeneity within the tumor, and is widely used in the local staging of rectal cancer and monitoring of treatment response.

[0004] As a product of the combination of computer technology and medicine, radiomics can extract a large number of features from medical images in a short time and convert grayscale changes visible to the naked eye into high-throughput data that can be analyzed by computers. Through analysis, many image features that cannot be found by conventional clinical radiologists when reading films can be obtained, thereby improving the accuracy of diagnosis, prognosis and prediction. Therefore, the present invention proposes a system and method for predicting the stability status of microsatellites in patients with rectal cancer. Summary of the invention

[0005] The embodiment of the present invention provides a system and method for predicting the microsatellite stability status of patients with rectal cancer to solve the above-mentioned technical problems in the prior art.

[0006] In order to have a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This summary is not intended to be a general review, nor is it intended to identify key / important components or to delineate the scope of protection of these embodiments. Its only purpose is to present some concepts in a simple form as a preface to the detailed description that follows.

[0007] According to a first aspect of an embodiment of the present invention, a system for predicting the microsatellite stability status of patients with rectal cancer is provided.

[0008] In one embodiment, the system for predicting microsatellite stability status of patients with rectal cancer comprises:

[0009] The training data set generation module is used to obtain the patient's electronic medical record data, postoperative pathology report data and MRI report data, and perform data preprocessing to obtain the training data set;

[0010] Microsatellite stability status assessment module, which is used to determine the expression status of the protein products of the mismatch repair system genes by immunohistochemical staining of postoperative paraffin-embedded specimens based on a standard procedure based on streptavidin-biotin-peroxidase;

[0011] An image acquisition and segmentation module is used to obtain a patient's T2-weighted imaging and diffusion-weighted imaging sequence images, and segment the tumor area on each continuous oblique axis T2-weighted imaging and diffusion-weighted imaging sequence image to obtain a three-dimensional volume map of the tumor area;

[0012] The radiomics feature extraction and screening module is used to extract the radiomics features of each tumor area in the three-dimensional volume image, and screen the extracted radiomics features based on the normal distribution method and the minimum absolute shrinkage and selection operator to obtain the key features for predicting the stability status of microsatellites;

[0013] A stability state prediction model building module is used to build and train a stability state prediction model using key features for predicting microsatellite stability states in a training data set;

[0014] The microsatellite stability status prediction module is used to predict the microsatellite stability status of rectal cancer patients based on the trained stability status prediction model combined with MRI images and clinical variables.

[0015] In one embodiment, the electronic medical record data includes data on gender, age, carcinoembryonic antigen, and tumor marker CA19-9;

[0016] The postoperative pathology report data include data on tumor differentiation degree and microsatellite stability status;

[0017] The MRI report data include imaging T stage, N stage, distance between the lower edge of the tumor and the anal verge, circumferential resection margin and extramuscular vascular invasion data.

[0018] In one embodiment, the training data set includes a training group and a validation group, and the ratio of the number of the training group to the number of the validation group is 7:3.

[0019] In one embodiment, the mismatch repair system gene protein products include MLH1, MSH2, MSH6, and PMS2.

[0020] In one embodiment, determining the expression status of a mismatch repair system gene protein product comprises:

[0021] When the expression of at least one gene protein product in the staining results is negative, the microsatellite is judged to be unstable; when the expression of all four gene protein products in the staining results is positive, the microsatellite is judged to be stable.

[0022] In one embodiment, during the segmentation process, non-invaded rectal wall, necrotic areas, and luminal contents are excluded, and intra- and inter-observer agreement is assessed using intraclass correlation coefficients.

[0023] In one embodiment, the categories of radiomic features include direct intensity, histogram of oriented gradients, gray-level co-occurrence matrix, gray-level run matrix, intensity histogram Gaussian fit, neighborhood intensity difference matrix, and shape-based features.

[0024] In one embodiment, before extracting the radiomics features, the magnetic resonance image is resampled based on B-spline interpolation, and the image grayscale value is normalized;

[0025] Before screening radiomic features, radiomic features with intraclass correlation coefficients less than or equal to the preset threshold were excluded, and each feature was z-score normalized and divided by the standard deviation.

[0026] In one embodiment, when the extracted radiomic features are screened based on the normal distribution method and the minimum absolute shrinkage and selection operator to obtain key features for predicting the stability state of microsatellites, a two-independent sample t-test or a Mann-Whitney U test is selected based on whether the radiomic features conform to the normal distribution; the radiomic features are preliminarily selected using the selected test method, and features with P values ​​greater than or equal to a preset threshold are excluded; redundant features are further excluded based on the minimum absolute shrinkage and selection operator to obtain key features for predicting the stability state of microsatellites.

[0027] According to a second aspect of an embodiment of the present invention, a method for predicting the microsatellite stability status of a rectal cancer patient is provided.

[0028] In one embodiment, the method for predicting the microsatellite stability status of a rectal cancer patient comprises:

[0029] Obtain the patient's electronic medical record data, postoperative pathology report data, and MRI report data, and perform data preprocessing to obtain a training data set;

[0030] Postoperative paraffin-embedded specimens were immunohistochemically stained based on a standard streptavidin-biotin-peroxidase procedure to determine the expression status of the protein products of the genes of the mismatch repair system;

[0031] Acquire the patient's T2-weighted imaging and diffusion-weighted imaging sequence images, and segment the tumor area on each of the continuous oblique axis T2-weighted imaging and diffusion-weighted imaging sequence images to obtain a three-dimensional volume map of the tumor area;

[0032] The radiomic features of each tumor region in the three-dimensional volume image were extracted, and the extracted radiomic features were screened based on the normal distribution method and the least absolute shrinkage and selection operator to obtain the key features for predicting the microsatellite stability status;

[0033] The key features for predicting the stability status of microsatellites in the training dataset were used to construct and train a stability status prediction model;

[0034] Based on the trained stability status prediction model, MRI images and clinical variables were combined to predict the microsatellite stability status of rectal cancer patients.

[0035] According to a third aspect of an embodiment of the present invention, a computer device is provided.

[0036] In some embodiments, the computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0037] According to a fourth aspect of embodiments of the present invention, a computer-readable storage medium is provided.

[0038] In one embodiment, the computer-readable storage medium stores a computer program, and the computer program implements the steps of the above method when executed by a processor.

[0039] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0040] The imaging omics of the present invention provides a non-invasive alternative method that can predict the stability of microsatellites through imaging data, reducing the pain and risk of patients. By establishing a prediction model that combines MRI image features and clinical variables, it can be used as a potential tool for non-invasive prediction of the stability of microsatellites in patients with stage II and III rectal cancer, which will be helpful in assisting clinical treatment decisions and achieving precision treatment.

[0041] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0043] Figure 1 It is a structural block diagram of a system for predicting microsatellite stability status of a rectal cancer patient according to an exemplary embodiment;

[0044] Figure 2 is a flow chart of a method for predicting microsatellite stability status of a rectal cancer patient according to an exemplary embodiment;

[0045] Figure 3 The figure is a schematic diagram showing the structure of a computer device according to an exemplary embodiment. DETAILED DESCRIPTION

[0046] The following description and accompanying drawings fully illustrate the specific embodiments of this article so that those skilled in the art can practice them. Parts and features of some embodiments may be included in or replace parts and features of other embodiments. The scope of the embodiments of this article includes the entire scope of the claims, as well as all available equivalents of the claims. Herein, the terms "first", "second", etc. are only used to distinguish one element from another, without requiring or implying any actual relationship or order between these elements. In fact, the first element can also be called the second element, and vice versa. Moreover, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that the structure, device or equipment including a series of elements includes not only those elements, but also other elements that are not explicitly listed, or also include elements inherent to such structure, device or equipment. In the absence of more restrictions, the elements defined by the sentence "including one..." do not exclude the existence of other identical elements in the structure, device or equipment including the elements. Each embodiment is described in a progressive manner herein, and each embodiment focuses on the differences from other embodiments, and the same and similar parts between the embodiments can be referred to each other.

[0047] The terms "longitudinal", "lateral", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc. in this document indicate the orientation or position relationship based on the orientation or position relationship shown in the drawings, and are only for the convenience of describing this document and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In the description of this document, unless otherwise specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a mechanical connection or an electrical connection, it can also be the internal communication of two elements, it can be a direct connection, or it can be an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0048] As used herein, the term "plurality" means two or more than two, unless otherwise specified.

[0049] In this document, the character " / " indicates that the preceding and following objects are in an "or" relationship. For example, A / B means: A or B.

[0050] In this article, the term "and / or" is a description of the association relationship between objects, indicating that three relationships may exist. For example, A and / or B means: A or B, or, A and B.

[0051] It should be understood that, although the various steps in the flow chart are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the figure may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

[0052] Each module in the device or system of the present application can be implemented in whole or in part by software, hardware, or a combination thereof. The above modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to the above modules.

[0053] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0054] Figure 1An embodiment of a system for predicting the microsatellite stability status of a rectal cancer patient according to the present invention is shown.

[0055] In this optional embodiment, the microsatellite stability status prediction system for rectal cancer patients includes:

[0056] The training data set generation module 101 is used to obtain the patient's electronic medical record data, postoperative pathology report data and MRI report data, and perform data preprocessing to obtain a training data set;

[0057] A microsatellite stability status assessment module 103 is used to perform immunohistochemical staining on postoperative paraffin-embedded specimens based on a standard procedure of streptavidin-biotin-peroxidase to determine the expression status of the protein products of the mismatch repair system genes;

[0058] The image acquisition and segmentation module 105 is used to obtain the patient's T2-weighted imaging and diffusion-weighted imaging sequence images, and segment the tumor area on each continuous oblique axis T2-weighted imaging and diffusion-weighted imaging sequence image to obtain a three-dimensional volume map of the tumor area;

[0059] The radiomics feature extraction and screening module 107 is used to extract the radiomics features of each tumor area in the three-dimensional volume image, and screen the extracted radiomics features based on the normal distribution method and the minimum absolute shrinkage and selection operator to obtain key features for predicting the stability status of microsatellites;

[0060] A stability state prediction model building module 109 is used to build and train a stability state prediction model using key features for predicting microsatellite stability states in a training data set;

[0061] The microsatellite stability state prediction module 111 is used to predict the microsatellite stability state of rectal cancer patients based on the trained stability state prediction model in combination with MRI images and clinical variables.

[0062] Figure 2 An embodiment of a method for predicting the microsatellite stability status of a rectal cancer patient according to the present invention is shown.

[0063] In this optional embodiment, the method for predicting the microsatellite stability status of a rectal cancer patient includes:

[0064] S201, obtaining the patient's electronic medical record data, postoperative pathology report data, and MRI report data, and performing data preprocessing to obtain a training data set;

[0065] S203, immunohistochemical staining of paraffin-embedded specimens after surgery based on standard streptavidin-biotin-peroxidase procedures to determine the expression status of the protein products of the mismatch repair system genes;

[0066] S205, acquiring a sequence of T2-weighted imaging and diffusion-weighted imaging images of the patient, and segmenting the tumor region on each of the continuous oblique axis T2-weighted imaging and diffusion-weighted imaging sequence images to obtain a three-dimensional volume map of the tumor region;

[0067] S207, extracting radiomic features of each tumor region in the three-dimensional volume image, and screening the extracted radiomic features based on the normal distribution method and the least absolute shrinkage and selection operator to obtain key features for predicting the microsatellite stability status;

[0068] S209, constructing and training a stability state prediction model using key features for predicting microsatellite stability states in the training data set;

[0069] S211. Based on the trained stability status prediction model, MRI images and clinical variables are combined to predict the microsatellite stability status of rectal cancer patients.

[0070] In order to facilitate understanding of the above technical solutions of the present invention, the above technical solutions of the present invention are further described from the perspective of architecture and principle as follows:

[0071] A method for predicting the microsatellite stability status of patients with stage II and stage III colorectal cancer is constructed to solve the problems of low prediction accuracy, lack of stability and individual specificity of existing clinical models. The specific technical solution is as follows:

[0072] (1) Study population: Consecutive cases who underwent surgery for rectal cancer in the Affiliated Hospital of XX University from February 2016 to October 2020 were analyzed. Inclusion criteria: (1) Postoperative pathological diagnosis of stage II-III rectal cancer; (2) High-resolution MRI images of the pelvis (rectum) within 2 weeks before surgical resection; (3) Known microsatellite stability status detected by immunohistochemistry. Exclusion criteria: (1) Poor MRI image quality, obvious artifacts, and inability to accurately mark the lesion outline; (2) Missing clinical pathological information; (3) Preoperative neoadjuvant chemoradiotherapy or other treatments. A total of 175 patients were included, including 119 males and 56 females, aged (63.9±9.4) years (range: 37-85 years).

[0073] (2) Data collection: Gender, age, carcinoembryonic antigen (normal reference value <3.4 μg / L), and tumor marker CA19-9 (normal reference value <39 U / ml) were obtained from the electronic medical records. Tumor differentiation and microsatellite stability status were obtained from the postoperative pathology report. Imaging T stage, N stage, distance from the lower edge of the tumor to the anal verge, circumferential resection margin, and extramyometrial vascular invasion were obtained from the MRI report.

[0074] (3) Training group and validation group: The included patients were randomly divided into a training group of 123 cases and a validation group of 52 cases in a ratio of 7:3 by a simple random method. There were 106 cases of microsatellite stability (MSS) and 17 cases of MSI (Microsatellite Instability) in the training group, and 46 cases of MSS and 6 cases of MSI in the validation group. There was no statistically significant difference in clinical data between the two groups.

[0075] (4) Assessment of microsatellite stability status: Based on the standard procedure of streptavidin-biotin-peroxidase, immunohistochemical staining was performed on the postoperative paraffin-embedded specimens to determine the expression status of four mismatch repair system gene protein products (MLH1, MSH2, MSH6, and PMS2). The staining results were judged by two pathologists. When there was a disagreement, a third pathology expert made the judgment. When at least one protein product was expressed negatively, it was judged as MSI; when all four protein products were expressed positively, it was judged as MSS. According to the immunohistochemical results, there were 152 cases in the MSS group and 23 cases in the MSI group. There were statistically significant differences in age, tumor differentiation degree, and the distance from the lower edge of the tumor to the anal verge between the MSS group and the MSI group (P < 0.05).

[0076] <0.05).

[0077] (5) Image acquisition and segmentation: All patients fasted for more than 4 hours before the examination and used enema to clean the intestines 1 hour before the examination. The examination equipment was a 3.0T Signa MRI with an eight-channel phased array coil. T2WI parameters: echo time 102ms, repetition time 3600ms, slice thickness 4mm, slice spacing 1mm, acquisition matrix 320×256, field of view 22cm 2 . DWI parameters: b value 1000s / mm 2 , echo time 94ms, repetition time 4200ms, layer 4mm, layer spacing 1mm, acquisition matrix 192×173, field of view 24cm 2The T2WI and DWI images of the included patients were exported from the hospital picture archiving and communication system. Two gastrointestinal radiologists with 10 years of clinical experience manually segmented the tumor area on each continuous oblique T2WI and DWI image using ITK-SNAP4.0.0 to obtain a three-dimensional volume map of the tumor area for feature extraction and quantification. During the segmentation process, the non-invaded rectal wall, necrotic area and intestinal contents were excluded. In order to ensure the consistency of manual segmentation and feature extraction, the intraclass correlation coefficient was used to evaluate the intra-observer and inter-observer consistency. That is, 30 cases were randomly selected, and the intraclass correlation coefficient between observers was calculated based on the extracted features. Subsequently, one of the radiologists re-annotated the above 30 randomly selected cases again after an interval of 1 month, and the intraclass correlation coefficient within the observer was calculated based on the extracted features. Features with intraclass correlation coefficients > 0.75 in both times were used for further analysis. Both radiologists were blinded to the microsatellite stability status of the patients.

[0078] The explanation and formula of the intraclass correlation coefficient are as follows:

[0079] The intraclass correlation coefficient (ICC) is used to evaluate the consistency and reliability of the measurement results of the same object by different observers (or the same observer at different time points). The calculation formula is as follows:

[0080]

[0081] Among them, MS B is the Between-group Mean Square, MS W is the within-group mean square, k is the number of measurements in each group, that is, the number of observers, which is 2 in this embodiment.

[0082] MS B The calculation formula is as follows:

[0083]

[0084] n is the number of objects, which is 30 in this embodiment; k is the number of observers, which is the same as above; is the mean of group i; is the grand mean of all observations.

[0085] MS W The calculation formula is as follows:

[0086]

[0087] X ijis the jth observation in the i-th group.

[0088] (6) Radiomic feature extraction and screening: Before extracting radiomic features, the magnetic resonance images were resampled using B-spline interpolation to ensure that all images had the same resolution. The image grayscale values ​​were normalized. The PyRadiomics software package was then used to extract radiomic features from each tumor region. The selected radiomic feature categories included direct intensity, oriented gradient histogram, grayscale co-occurrence matrix, grayscale run matrix, intensity histogram Gaussian fit, neighborhood intensity difference matrix, and shape-based features. Subsequently, in order to screen out the features most relevant to the microsatellite stability status of rectal cancer and reduce the risk of model overfitting, radiomic features with intraclass correlation coefficients ≤ 0.75 were first excluded, and each feature was z-score normalized, that is, each feature vector was subtracted from the mean and then divided by the standard deviation. The subsequent radiomic feature screening process was performed in the training set. Depending on whether the feature conformed to the normal distribution, a two-independent sample t-test or Mann-Whitney U test was performed as a preliminary selection, and features with a P value ≥ 0.05 were excluded. Afterwards, the least absolute shrinkage and selection operator (LASSO) was used to further eliminate redundant features and obtain the most significant features for predicting microsatellite stability status in the training group. LASSO is a commonly used linear regression method that introduces L1 regularization to achieve feature selection and model simplification. Previously, the patients were randomly divided into training and test sets, and the extracted radiomic features were z-score normalized. Next, the optimal α value was selected by ten-fold cross-validation, and the LASSO model was trained using the optimal α value. The selected features were determined by the coefficients of the LASSO model, that is, the features with non-zero coefficients were the selected features. In addition, the artificial minority class oversampling method was used to deal with the sample size imbalance between MSI and MSS. The results showed that the intra-group correlation coefficients of all extracted radiomic features were >0.75, with good intra-observer and inter-observer reproducibility. 71 T2WI-related features and 71 DWI-related features were extracted from the MRI images of patients in the training group. For the imaging features, six imaging features were finally screened out after intra-observer and inter-observer consistency tests, two independent sample t-tests or Mann-Whitney U tests and LASSO regression, including four T2WI-related features (surface volume ratio, cluster prominence, idmn, zone entropy) and two DWI-related features (mesh volume, gray level non-uniformity).For clinical characteristics, two independent sample t-tests or Mann-Whitney U tests were performed on the clinical data of patients in the MSI group and the MSS group. The results showed that there were statistical differences in three clinical characteristics (age, tumor differentiation degree, and distance from the lower edge of the tumor to the anal verge) between the two groups.

[0089] (7) Model establishment and verification: The models established in this embodiment are all based on the support vector machine classifier, which has many advantages in establishing prediction models and is particularly suitable for classification tasks involving high-dimensional, complex, and small sample data sets. In order to explore whether clinical features can provide additional information for predicting microsatellite stability status, the above three clinical features were z-score normalized and a clinical model was established using a support vector machine classifier. Based on the above four T2WI-related features and two DWI-related features, a T2WI model and a DWI model were established respectively. The T2WI+DWI model was established by combining the six radiomics features. In addition, the clinical-radiological model was constructed by combining the six radiomics features and the three clinical features. The results showed that in the training group, the AUCs of the DWI model, the T2WI model, and the T2WI+DWI model were 0.81, 0.78, and 0.92, respectively; in the validation group, the AUCs of the above three models were 0.71, 0.65, and 0.82, respectively. The AUC of the T2WI+DWI model was higher than that of the other three models in both the training and validation groups. In the training group, the AUC of the clinical-radiomics model was 0.95, which was higher than that of the other models; in the validation group, the AUC of the clinical-radiomics model was 0.81, which was comparable to that of the T2WI+DWI model (Z=-0.21, P=0.83) and higher than that of the clinical model (Z=0.71, P=0.04).

[0090] The present invention can non-invasively extract a large number of imaging omics features from MRI images of patients with rectal cancer based on imaging omics, which are used to evaluate the microsatellite stability status of patients with rectal cancer. The microsatellite stability status is closely related to the aggressiveness of cancer and the prognosis of patients. Predicting the microsatellite stability status through imaging omics can provide additional diagnostic information to help doctors more accurately assess the condition and prognosis. In addition, the microsatellite stability status of rectal cancer has an important impact on the choice of treatment strategy. Patients with MSI status are usually more sensitive to immunotherapy (such as PD-1 / PD-L1 inhibitors), while patients with MSS status may be more sensitive to chemotherapy. Imaging omics can assist in predicting the microsatellite stability status of patients, thereby helping to formulate personalized treatment plans and improve the treatment effect. Traditional microsatellite stability detection usually requires obtaining tumor samples through tissue biopsy and then performing molecular biological testing. This method is invasive and may cause complications. Imaging omics provides a non-invasive alternative method that can predict the microsatellite stability status through imaging data, reducing the pain and risk of patients.

[0091] In summary, the present invention has established a prediction model combining MRI image features and clinical variables, which can be used as a potential tool for non-invasive prediction of the microsatellite stability status of patients with stage II and III rectal cancer, which will be helpful in assisting clinical treatment decision-making and achieving precision treatment.

[0092] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 3 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store static information and dynamic information data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the steps in the above method embodiment are implemented.

[0093] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0094] In addition, the present invention also provides a computer device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiment when executing the computer program.

[0095] In addition, the present invention further provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps in the above method embodiment are implemented.

[0096] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0097] The present invention is not limited to the structures which have been described above and shown in the drawings, and various modifications and changes may be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. A system for predicting the microsatellite stability status of patients with rectal cancer, characterized in that: include: The training data set generation module is used to obtain the patient's electronic medical record data, postoperative pathology report data and MRI report data, and perform data preprocessing to obtain the training data set; Microsatellite stability status assessment module, which is used to determine the expression status of the protein products of the mismatch repair system genes by immunohistochemical staining of postoperative paraffin-embedded specimens based on a standard procedure based on streptavidin-biotin-peroxidase; An image acquisition and segmentation module is used to obtain a patient's T2-weighted imaging and diffusion-weighted imaging sequence images, and segment the tumor area on each continuous oblique axis T2-weighted imaging and diffusion-weighted imaging sequence image to obtain a three-dimensional volume map of the tumor area; The radiomics feature extraction and screening module is used to extract the radiomics features of each tumor area in the three-dimensional volume image, and screen the extracted radiomics features based on the normal distribution method and the minimum absolute shrinkage and selection operator to obtain the key features for predicting the stability status of microsatellites; A stability state prediction model building module is used to build and train a stability state prediction model using key features for predicting microsatellite stability states in a training data set; The microsatellite stability status prediction module is used to predict the microsatellite stability status of rectal cancer patients based on the trained stability status prediction model combined with MRI images and clinical variables.

2. The microsatellite stability status prediction system for rectal cancer patients according to claim 1, characterized in that: The electronic medical record data includes data on gender, age, carcinoembryonic antigen and tumor marker CA19-9; The postoperative pathology report data include data on tumor differentiation degree and microsatellite stability status; The MRI report data include imaging T stage, N stage, distance between the lower edge of the tumor and the anal verge, circumferential resection margin and extramuscular vascular invasion data.

3. The microsatellite stability status prediction system for rectal cancer patients according to claim 1, characterized in that: The training data set includes a training group and a validation group, and the ratio of the number of the training group to the number of the validation group is 7:

3.

4. The microsatellite stability status prediction system for rectal cancer patients according to claim 1, characterized in that: The mismatch repair system gene protein products include four types: MLH1, MSH2, MSH6, and PMS2.

5. The microsatellite stability status prediction system for rectal cancer patients according to claim 1, characterized in that: Determining the expression status of the protein products of the mismatch repair system genes includes: When the expression of at least one gene protein product in the staining results is negative, the microsatellite is judged to be unstable; when the expression of all four gene protein products in the staining results is positive, the microsatellite is judged to be stable.

6. The microsatellite stability status prediction system for rectal cancer patients according to claim 1, characterized in that: During segmentation, non-invaded rectal wall, necrotic areas, and luminal contents were excluded, and intra- and inter-observer agreement was assessed using the intraclass correlation coefficient.

7. The system for predicting the microsatellite stability status of patients with rectal cancer according to claim 1, characterized in that: The categories of radiomic features include direct intensity, histogram of oriented gradients, gray-level co-occurrence matrix, gray-level run matrix, intensity histogram Gaussian fit, neighborhood intensity difference matrix, and shape-based features.

8. The system for predicting microsatellite stability status of patients with rectal cancer according to claim 1, characterized in that: Before extracting radiomic features, the magnetic resonance images were resampled based on B-spline interpolation, and the image grayscale values ​​were normalized; Before screening radiomic features, radiomic features with intraclass correlation coefficients less than or equal to the preset threshold were excluded, and each feature was z-score normalized and divided by the standard deviation.

9. The system for predicting the microsatellite stability status of patients with rectal cancer according to claim 1, characterized in that: When the extracted radiomic features are screened based on the normal distribution method and the least absolute shrinkage and selection operator to obtain the key features for predicting the stability status of microsatellites, two independent sample t-tests or Mann-Whitney U test methods are selected based on whether the radiomic features conform to the normal distribution; the radiomic features are preliminarily selected using the selected test method, and the features with P values ​​greater than or equal to the preset threshold are excluded; Redundant features were eliminated based on the least absolute shrinkage and selection operators to obtain key features for predicting the stability status of microsatellites.

10. A method for predicting the microsatellite stability status of patients with rectal cancer, characterized in that: include: Obtain the patient's electronic medical record data, postoperative pathology report data, and MRI report data, and perform data preprocessing to obtain a training data set; Postoperative paraffin-embedded specimens were immunohistochemically stained based on a standard streptavidin-biotin-peroxidase procedure to determine the expression status of the protein products of the genes of the mismatch repair system; Acquire the patient's T2-weighted imaging and diffusion-weighted imaging sequence images, and segment the tumor area on each of the continuous oblique axis T2-weighted imaging and diffusion-weighted imaging sequence images to obtain a three-dimensional volume map of the tumor area; The radiomic features of each tumor region in the three-dimensional volume image were extracted, and the extracted radiomic features were screened based on the normal distribution method and the least absolute shrinkage and selection operator to obtain the key features for predicting the microsatellite stability status; The key features for predicting the stability status of microsatellites in the training dataset were used to construct and train a stability status prediction model; Based on the trained stability status prediction model, MRI images and clinical variables were combined to predict the microsatellite stability status of rectal cancer patients.

Images