Stomach tissue pathological risk assessment method and device, electronic equipment and medium

Abstract

The embodiment of the invention relates to the technical field of medical detection, and provides a stomach tissue pathology risk assessment method and device, electronic equipment and a medium, and the method comprises the following steps: obtaining target Raman spectrum detection data of a stomach tissue sample of a current patient; inputting the detection data into the stomach tissue pathology risk assessment model to obtain an output result, and further determining a stomach tissue pathology risk assessment result of the current patient; the method comprises the following steps: acquiring Raman spectrum detection data corresponding to a plurality of stomach tissue samples by the stomach tissue pathology risk assessment model; dividing the detection data into a training set and a verification set; the initial model is trained based on the training set, and then the verification set is evaluated; and when the effectiveness of the assessment result meets a preset condition, obtaining a stomach tissue pathology risk assessment model. Therefore, the Raman spectrum and the artificial intelligence technology are combined for evaluating the pathological risk of the stomach tissue, rapid and accurate detection of the cancer, non-cancer and precancerous pathological stages of the stomach tissue sample is realized, and a diagnosis reference is provided for doctors.

Description

Technical Field

[0001] The present invention relates to the field of medical detection technology, and in particular to a gastric tissue pathology risk assessment method, device, electronic equipment and medium. Background Art

[0002] Gastric cancer is currently the fifth most common cancer in the world. According to statistics, the 5-year survival rate of gastric cancer is about 25%-30%. Early diagnosis and treatment can effectively improve the survival rate of gastric cancer. The current gold standard for clinical diagnosis of gastric cancer is endoscopic biopsy, which is easily affected by patient compliance and operating techniques. In addition, tissue biopsy requires H&E or immunohistochemical staining, which takes a long time to wait and requires experienced pathologists to interpret the results, which also cannot avoid the subjectivity of the pathologist's diagnosis.

[0003] Therefore, how to quickly, accurately and label-freely perform gastric cancer tissue pathology detection has become an urgent problem to be solved. Summary of the invention

[0004] The present invention provides a gastric tissue pathology risk assessment method, device, electronic device and medium, which are used to solve the defect in the prior art that endoscopic biopsy cannot be accurately and quickly detected due to the influence of patient compliance and operating technology, and realize rapid and accurate detection of cancer, non-cancerous and precancerous pathological stages of gastric tissue samples, providing a diagnostic reference for doctors.

[0005] The present invention provides a gastric tissue pathology risk assessment method, comprising: Acquire target Raman spectrum detection data of the gastric tissue sample to be evaluated of the current patient; Inputting the target Raman spectrum detection data into a gastric tissue pathology risk assessment model to obtain an output result of the gastric tissue pathology risk assessment model; Determine the gastric tissue pathology risk assessment result of the current patient based on the output result; The gastric tissue pathology risk assessment model is trained based on the following steps: Acquire Raman spectrum detection data corresponding to a plurality of gastric tissue samples, wherein each piece of Raman spectrum detection data corresponds to a target gastric disease symptom; Dividing the Raman spectrum detection data into a model training set and a model verification set; Training the initial model based on the model training set to obtain a trained initial model; Evaluate the model validation set based on the trained initial model to obtain an evaluation result; When the validity of the evaluation result meets the preset conditions, it is determined that the initial model training is completed, and the trained initial model is used as the gastric tissue pathology risk evaluation model.

[0006] In a possible implementation, the method further includes: The stomach tissue samples include at least stomach cancer tissue samples and non-gastric cancer tissue samples, and the Raman spectrum detection data include first Raman spectrum detection data and second Raman spectrum detection data; Raman spectrum detection is performed on multiple gastric cancer tissue samples and non-gastric cancer tissue samples under preset Raman spectrum detection conditions to obtain the first Raman spectrum detection data corresponding to the multiple gastric cancer tissue samples and the second Raman spectrum detection data corresponding to the non-gastric cancer tissue samples.

[0007] In a possible implementation, the method further includes: The first Raman spectrum detection data and the second Raman spectrum detection data are respectively divided into a model training set and a model verification set according to a preset ratio, the model training set contains the first Raman spectrum detection data and the second Raman spectrum detection data in a preset ratio, and the model verification set contains the first Raman spectrum detection data and the second Raman spectrum detection data in a preset ratio.

[0008] In a possible implementation, the method further includes: The Raman spectrum detection data in the model training set and the target gastric disease symptom corresponding to each piece of Raman spectrum detection data are input into the initial model, and the initial model is trained to obtain a trained initial model.

[0009] In a possible implementation, the method further includes: Inputting the model verification set into the trained initial model so that the trained initial model evaluates the model verification set to obtain an evaluation result; Determine the validity of the evaluation results.

[0010] In a possible implementation, the method further includes: If the validity of the evaluation result does not meet the preset condition, re-executing the step of training the initial model based on the model training set to obtain the trained initial model; If the validity of the evaluation result meets the preset conditions, it is determined that the initial model training is completed, and the trained initial model is used as the gastric tissue pathology risk evaluation model.

[0011] The present invention also provides a gastric tissue pathology risk assessment device, comprising: An acquisition module, used for acquiring target Raman spectrum detection data of a gastric tissue sample to be evaluated of a current patient; An evaluation module, used for inputting the target Raman spectrum detection data into a gastric tissue pathology risk evaluation model to obtain an output result of the gastric tissue pathology risk evaluation model; A determination module, configured to determine a gastric tissue pathology risk assessment result of the current patient based on the output result; A model training module is used to obtain Raman spectral detection data corresponding to multiple gastric tissue samples, wherein each piece of Raman spectral detection data corresponds to a target gastric disease symptom; divide the Raman spectral detection data into a model training set and a model verification set; train the initial model based on the model training set to obtain a trained initial model; evaluate the model verification set based on the trained initial model to obtain an evaluation result; when the validity of the evaluation result meets a preset condition, determine that the initial model training is completed, and use the trained initial model as a gastric tissue pathology risk assessment model.

[0012] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, a gastric tissue pathology risk assessment method as described above is implemented.

[0013] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the gastric tissue pathology risk assessment method as described in any one of the above is implemented.

[0014] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the gastric tissue pathology risk assessment method as described in any one of the above is implemented.

[0015] The gastric tissue pathology risk assessment method, device, electronic device and medium provided by the present invention obtain the target Raman spectrum detection data of the gastric tissue sample to be evaluated of the current patient; input the target Raman spectrum detection data into the gastric tissue pathology risk assessment model to obtain the output result of the gastric tissue pathology risk assessment model; determine the gastric tissue pathology risk assessment result of the current patient based on the output result; wherein the gastric tissue pathology risk assessment model is trained based on the following steps: obtain Raman spectrum detection data corresponding to a plurality of gastric tissue samples, wherein each Raman spectrum detection data corresponds to a target gastric disease symptom; divide the Raman spectrum detection data into a model training set and a model verification set; train the initial model based on the model training set to obtain the trained initial model; evaluate the model verification set based on the trained initial model to obtain an evaluation result; when the validity of the evaluation result meets the preset conditions, determine that the initial model training is completed, and use the trained initial model as the gastric tissue pathology risk assessment model. Compared with the defect that gastroscopic biopsy in the prior art cannot be accurately and quickly detected due to the influence of patient compliance and operation technology. Therefore, combining Raman spectroscopy with artificial intelligence technology for the precise assessment of gastric tissue pathology risk can achieve rapid and accurate detection of gastric tissue samples at the cancer, non-cancerous, and precancerous pathology stages, providing doctors with a diagnostic reference. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0017] Figure 1 It is a schematic diagram of a process of gastric tissue pathology risk assessment method provided by the present invention.

[0018] Figure 2 It is a flow chart of a training method for a gastric tissue pathology risk assessment model provided by the present invention.

[0019] Figure 3 It is a structural schematic diagram of a gastric tissue pathology risk assessment device provided by the present invention.

[0020] Figure 4 It is the average Raman spectrum of the gastric cancer samples and non-cancerous samples provided by the present invention.

[0021] Figure 5 t-SNE diagram of gastric cancer samples and non-cancerous samples provided by the present invention.

[0022] Figure 63 is a ROC curve diagram of gastric cancer samples and non-cancerous samples provided by the present invention.

[0023] Figure 7 It is the average Raman spectrum of five pathological types of gastric tissue provided by the present invention.

[0024] Figure 8 It is a t-SNE diagram of five pathological types of gastric tissue provided by the present invention.

[0025] Fig. 9 It is a ROC curve diagram of five pathological types of gastric tissue provided by the present invention.

[0026] Fig.10 It is the average Raman spectrum of the Helicobacter pylori positive and Helicobacter pylori negative samples provided by the present invention.

[0027] Fig.11 It is a t-SNE diagram of Helicobacter pylori positive and Helicobacter pylori negative samples provided by the present invention.

[0028] Fig.12 It is a ROC curve diagram of Helicobacter pylori positive and Helicobacter pylori negative samples provided by the present invention.

[0029] Fig.13 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0031] To facilitate understanding of the embodiments of the present invention, specific embodiments will be further explained below in conjunction with the accompanying drawings. The embodiments do not constitute a limitation on the embodiments of the present invention.

[0032] Figure 1 : is a schematic diagram of a gastric tissue pathology risk assessment method provided by the present invention, such as Figure 1 As shown, the method includes the following: S11. Obtain target Raman spectrum detection data of the gastric tissue sample to be evaluated of the current patient.

[0033] Under preset Raman spectral detection conditions, Raman spectral detection is performed on the gastric tissue slice to be evaluated of the current patient to obtain target Raman spectral detection data of the gastric tissue sample to be evaluated of the current patient.

[0034] S12, inputting the target Raman spectrum detection data into a gastric tissue pathology risk assessment model to obtain an output result of the gastric tissue pathology risk assessment model.

[0035] The target Raman spectrum detection data is input into the gastric tissue pathology risk assessment model, so that the gastric tissue pathology risk assessment model makes an estimate for the target Raman spectrum detection data, and obtains the estimated gastric disease symptoms output by the model.

[0036] S13. Determine a gastric tissue pathology risk assessment result of the current patient based on the output result.

[0037] The gastric disease symptoms of the current patient's gastric tissue pathology are judged based on the model output results, and the judgment results can assist doctors in making rapid clinical diagnosis.

[0038] The gastric tissue pathology risk assessment method provided by the present invention obtains the target Raman spectrum detection data of the gastric tissue sample to be evaluated of the current patient; inputs the target Raman spectrum detection data into the gastric tissue pathology risk assessment model to obtain the model output result; and determines the gastric tissue pathology risk assessment result of the current patient based on the model output result. Compared with the defect of gastroscopy biopsy in the prior art that cannot be accurately and quickly detected due to the influence of patient compliance and operation technology. By this method, Raman spectroscopy and artificial intelligence technology are combined for the accurate assessment of gastric tissue pathology risk, which can realize the rapid and accurate detection of gastric tissue samples in cancer, non-cancerous and precancerous pathology stages, and provide a diagnostic reference for doctors.

[0039] Figure 2 is a flow chart of a method for training a gastric tissue pathology risk assessment model provided by the present invention, such as Figure 2 As shown, the method includes the following: S21. Obtain Raman spectrum detection data corresponding to a plurality of gastric tissue samples.

[0040] In the embodiment of the present invention, a plurality of gastric mucosal tissue samples (including gastric cancer tissue samples and non-gastric cancer tissue samples) are first obtained for tissue sectioning. The tissue sections include frozen sections and paraffin sections, wherein the paraffin sections need to be dewaxed. There is no size limit for the thickness and size of the tissue sections. The tissue sections can be placed on various slides or substrates, including but not limited to glass slides, silicon wafers, aluminum sheets, CaF2 substrates, etc.

[0041] Under the preset Raman spectrum detection conditions, Raman spectrum detection is performed on multiple gastric cancer tissue samples and non-gastric cancer tissue samples to obtain first Raman spectrum detection data corresponding to the multiple gastric cancer tissue samples and second Raman spectrum detection data corresponding to the non-gastric cancer tissue samples. Each Raman spectrum detection data corresponds to a target gastric disease symptom, such as superficial gastritis, atrophic gastritis, intestinal metaplasia, dysplasia, early cancer, Helicobacter pylori positive, Helicobacter pylori negative, etc.

[0042] In the embodiment of the present invention, the Raman spectrometer includes but is not limited to a handheld or portable Raman spectrometer, a microscopic confocal Raman spectrometer and a stimulated Raman spectrometer.

[0043] Raman spectroscopy detection conditions include selecting laser wavelength, setting integration time, laser power, number of acquisitions, grating size, detection wavenumber range, etc.

[0044] Furthermore, the Raman spectrum detection data is preprocessed, including but not limited to removal of abnormal curves, baseline subtraction, cosmic ray subtraction, fluorescence background subtraction, Raman spectrum smoothing and noise reduction processing.

[0045] S22, dividing the Raman spectrum detection data into a model training set and a model verification set.

[0046] The first Raman spectrum detection data and the second Raman spectrum detection data after the above preprocessing are divided into a model training set and a model verification set respectively. The model training set contains the first Raman spectrum detection data and the second Raman spectrum detection data in a preset ratio, and the model verification set contains the first Raman spectrum detection data and the second Raman spectrum detection data in a preset ratio. The ratio can be determined according to actual needs, such as 2:1, 3:1, 4:1 or 5:1.

[0047] S23. Train the initial model based on the model training set to obtain a trained initial model.

[0048] The Raman spectrum detection data in the model training set and the target gastric disease symptoms corresponding to each Raman spectrum detection data are input into the initial model, and the initial model is trained to obtain a trained initial model.

[0049] The initial model must use machine learning or artificial intelligence algorithms, including but not limited to Gated Recurrent Unit (GRU), Support Vector Machine (SVM), Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA), K-Nearest Neighbor (KNN), Naive Bayes model (NB), and Convolutional Neural Networks (CNN).

[0050] S24. Evaluate the model verification set based on the trained initial model to obtain an evaluation result.

[0051] The model validation set is input into the trained initial model so that the trained initial model evaluates the model validation set to obtain an evaluation result; and the validity of the evaluation result is judged. The validity of the evaluation result is judged by: the model outputs the estimated gastric disease symptoms corresponding to each Raman spectrum detection data in the model validation set, and compares the actual gastric disease symptoms corresponding to each Raman spectrum detection data in the model validation set. If the estimated accuracy reaches a preset threshold (for example, 98%), the evaluation result can be determined to be valid; if the estimated accuracy does not reach the preset threshold, the evaluation result can be determined to be invalid.

[0052] S25. When the validity of the evaluation result meets a preset condition, it is determined that the initial model training is completed, and the trained initial model is used as a gastric tissue pathology risk evaluation model.

[0053] When the evaluation result is valid, it is determined that the initial model training is completed, and the trained initial model is used as the gastric tissue pathology risk assessment model.

[0054] Optionally, when the evaluation result is invalid, adjust the parameters of the model and continue to train the model using the model training set.

[0055] Specific embodiments are given below: This example is used to illustrate the use of Raman spectroscopy combined with machine learning algorithms for the diagnosis of cancer and non-cancerous gastric tissues: Gastric tissue samples were divided into two categories: cancer and non-cancerous. 2468 Raman spectra of gastric cancer samples and 542 Raman spectra of non-cancerous samples were collected. After Raman spectroscopy preprocessing, 2000 Raman spectra of gastric cancer samples and 500 Raman spectra of non-cancerous samples were retained for model training. 20% of the spectra were randomly selected as the validation set. The average Raman spectra of gastric cancer samples and non-cancerous samples are shown in Figure 4 By displaying the data using the t-SNE dimensionality reduction method, we can see that gastric cancer samples and non-cancer samples can be well classified ( Figure 5 ). During the GRU model training process, through the ROC curve, we can see that the AUC is 0.98, indicating that the GRU model has good classification performance ( Figure 6 ). The classification accuracy was calculated to be 95%, the sensitivity was 92%, and the specificity was 92%.

[0056] This example is used to illustrate the use of Raman spectroscopy combined with machine learning algorithms for gastric tissue pathology diagnosis: Gastric tissue samples were divided into five categories: superficial gastritis, atrophic gastritis, intestinal metaplasia, dysplasia, and early cancer. 712 Raman spectra of superficial gastritis samples, 603 Raman spectra of atrophic gastritis samples, 816 Raman spectra of intestinal metaplasia samples, 517 Raman spectra of dysplasia samples, and 548 Raman spectra of early cancer samples were collected. After Raman spectral preprocessing, 500 Raman spectra of each type of sample were retained for model training. 20% of the spectra were randomly selected as the validation set. The average Raman spectra of gastric tissue samples of the five pathological types are shown in Figure 2. Figure 7 . By displaying the data using the t-SNE dimensionality reduction method, we can see that these five types of pathological samples can be well classified ( Figure 8 ). During the GRU model training process, through the ROC curve, we can see that the AUCs of superficial gastritis, atrophic gastritis, intestinal metaplasia, dysplasia, and early cancer are 0.98 respectively, indicating that the GRU model has good classification performance ( Fig. 9 ). The classification accuracy was calculated to be 89%, the sensitivity was 89%, and the specificity was 97%.

[0057] This example is used to illustrate the use of Raman spectroscopy combined with machine learning algorithms for the diagnosis of gastric tissue pathology of cancer and non-cancerous tissues: Gastric tissue samples were divided into two categories: Helicobacter pylori positive and Helicobacter pylori negative. 2131 Raman spectra of Helicobacter pylori positive samples and 1065 Raman spectra of Helicobacter pylori negative samples were collected. After Raman spectroscopy preprocessing, 1500 Raman spectra of Helicobacter pylori positive samples and 1000 Raman spectra of Helicobacter pylori negative samples were retained for model training. 20% of the spectra were randomly selected as the validation set. The average Raman spectra of Helicobacter pylori positive samples and Helicobacter pylori negative samples are shown in Figure 2. Fig.10. By displaying the data using the t-SNE dimensionality reduction method, we can see that Helicobacter pylori positive samples and Helicobacter pylori negative samples can be well classified ( Fig.11 ). During the GRU model training process, through the ROC curve, we can see that the AUC is 0.99, indicating that the GRU model has good classification performance ( Fig.12 ). The classification accuracy, sensitivity and specificity were calculated to be 96%, 96% and 96%, respectively.

[0058] After testing, the gastric tissue pathology risk assessment model constructed as above has an accuracy rate greater than 80%, a sensitivity greater than 80%, and a specificity greater than 80%. Preferably, the accuracy rate is greater than 85%, the sensitivity is greater than 85%, and the specificity is greater than 95%.

[0059] The gastric tissue pathology risk assessment model provided by the present invention can simultaneously detect cancer and non-cancerous, different precancerous pathological stages and Helicobacter pylori, has the advantage of simultaneous detection of multiple indicators, and can save time and cost.

[0060] The training method of the gastric tissue pathology risk assessment model provided by the present invention is to obtain Raman spectrum detection data corresponding to multiple gastric tissue samples, wherein each Raman spectrum detection data corresponds to the target gastric disease symptom; the Raman spectrum detection data is divided into a model training set and a model verification set; the initial model is trained based on the model training set to obtain the trained initial model; the model verification set is evaluated based on the trained initial model to obtain an evaluation result; when the validity of the evaluation result meets the preset conditions, it is determined that the initial model training is completed, and the trained initial model is used as the gastric tissue pathology risk assessment model. Compared with the defect that gastroscopy biopsy in the prior art cannot be accurately and quickly detected due to the influence of patient compliance and operation technology. By this method, Raman spectroscopy and artificial intelligence technology are combined for the accurate assessment of gastric tissue pathology risk, which can realize the rapid and accurate detection of gastric tissue samples in cancer, non-cancerous and precancerous pathological stages, and provide a diagnostic reference for doctors.

[0061] The gastric tissue pathology risk assessment device provided by the present invention is described below. The gastric tissue pathology risk assessment device described below and the gastric tissue pathology risk assessment method described above can be referenced to each other.

[0062] Figure 3 Schematic diagram of a gastric tissue pathology risk assessment device provided by the present invention. Figure 3 As shown, the device includes the following: An acquisition module 301 is used to acquire target Raman spectrum detection data of a gastric tissue sample to be evaluated of a current patient; An evaluation module 302 is used to input the target Raman spectrum detection data into a gastric tissue pathology risk evaluation model to obtain an output result of the gastric tissue pathology risk evaluation model; A determination module 303 is used to determine the gastric tissue pathology risk assessment result of the current patient based on the output result; The model training module 304 is used to obtain Raman spectral detection data corresponding to multiple gastric tissue samples, wherein each Raman spectral detection data corresponds to a target gastric disease symptom; divide the Raman spectral detection data into a model training set and a model verification set; train the initial model based on the model training set to obtain a trained initial model; evaluate the model verification set based on the trained initial model to obtain an evaluation result; when the validity of the evaluation result meets the preset conditions, determine that the initial model training is completed, and use the trained initial model as a gastric tissue pathology risk assessment model.

[0063] Fig.13 An example of a physical structure diagram of an electronic device is shown in FIG. Fig.13 As shown, the electronic device may include: a processor (processor) 810, a communication interface (Communications Interface) 820, a memory (memory) 830 and a communication bus 840, wherein the processor 810, the communication interface 820, and the memory 830 communicate with each other through the communication bus 840. The processor 810 can call the logic instructions in the memory 830 to execute the gastric tissue pathology risk assessment method, which includes: obtaining target Raman spectrum detection data of the gastric tissue sample to be evaluated of the current patient; inputting the target Raman spectrum detection data into the gastric tissue pathology risk assessment model to obtain the output result of the gastric tissue pathology risk assessment model; determining the gastric tissue pathology risk assessment result of the current patient based on the output result; wherein the gastric tissue pathology risk assessment model is trained based on the following steps: obtaining Raman spectrum detection data corresponding to multiple gastric tissue samples, wherein each Raman spectrum detection data corresponds to a target gastric disease symptom; dividing the Raman spectrum detection data into a model training set and a model verification set; training the initial model based on the model training set to obtain a trained initial model; evaluating the model verification set based on the trained initial model to obtain an evaluation result; when the validity of the evaluation result meets the preset conditions, determining that the initial model training is completed, and using the trained initial model as the gastric tissue pathology risk assessment model.

[0064] In addition, the logic instructions in the above-mentioned memory 830 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0065] On the other hand, the present invention also provides a computer program product, which includes a computer program, which can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the gastric tissue pathology risk assessment method provided by the above-mentioned methods, which includes: obtaining target Raman spectrum detection data of the gastric tissue sample to be evaluated of the current patient; inputting the target Raman spectrum detection data into the gastric tissue pathology risk assessment model to obtain the output result of the gastric tissue pathology risk assessment model; determining the gastric tissue pathology risk assessment result of the current patient based on the output result; wherein the gastric tissue pathology risk assessment model is trained based on the following steps: obtaining Raman spectrum detection data corresponding to a plurality of gastric tissue samples, wherein each Raman spectrum detection data corresponds to a target gastric disease symptom; dividing the Raman spectrum detection data into a model training set and a model verification set; training the initial model based on the model training set to obtain a trained initial model; evaluating the model verification set based on the trained initial model to obtain an evaluation result; when the validity of the evaluation result meets the preset conditions, determining that the initial model training is completed, and using the trained initial model as the gastric tissue pathology risk assessment model.

[0066] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the gastric tissue pathology risk assessment method provided by the above-mentioned methods, the method comprising: obtaining target Raman spectrum detection data of the gastric tissue sample to be evaluated of the current patient; inputting the target Raman spectrum detection data into a gastric tissue pathology risk assessment model to obtain an output result of the gastric tissue pathology risk assessment model; determining the gastric tissue pathology risk assessment result of the current patient based on the output result; wherein the gastric tissue pathology risk assessment model is trained based on the following steps: obtaining Raman spectrum detection data corresponding to a plurality of gastric tissue samples, respectively, wherein each Raman spectrum detection data corresponds to a target gastric disease symptom; dividing the Raman spectrum detection data into a model training set and a model verification set; training an initial model based on the model training set to obtain a trained initial model; evaluating the model verification set based on the trained initial model to obtain an evaluation result; when the validity of the evaluation result meets a preset condition, determining that the training of the initial model is completed, and using the trained initial model as the gastric tissue pathology risk assessment model.

[0067] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0068] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical 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 embodiment or some parts of the embodiments.

[0069] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A gastric tissue pathology risk assessment method, characterized in that: include: Acquire target Raman spectrum detection data of the gastric tissue sample to be evaluated of the current patient; Inputting the target Raman spectrum detection data into a gastric tissue pathology risk assessment model to obtain an output result of the gastric tissue pathology risk assessment model; Determine the gastric tissue pathology risk assessment result of the current patient based on the output result; The gastric tissue pathology risk assessment model is trained based on the following steps: Acquire Raman spectrum detection data corresponding to a plurality of gastric tissue samples, wherein each piece of Raman spectrum detection data corresponds to a target gastric disease symptom; Dividing the Raman spectrum detection data into a model training set and a model verification set; Training the initial model based on the model training set to obtain a trained initial model; Evaluate the model validation set based on the trained initial model to obtain an evaluation result; When the validity of the evaluation result meets the preset conditions, it is determined that the initial model training is completed, and the trained initial model is used as the gastric tissue pathology risk evaluation model.

2. The method according to claim 1, characterized in that The stomach tissue samples include at least stomach cancer tissue samples and non-gastric cancer tissue samples, and the Raman spectrum detection data include first Raman spectrum detection data and second Raman spectrum detection data; The step of obtaining Raman spectroscopy detection data corresponding to a plurality of gastric tissue samples includes: Raman spectrum detection is performed on multiple gastric cancer tissue samples and non-gastric cancer tissue samples under preset Raman spectrum detection conditions to obtain the first Raman spectrum detection data corresponding to the multiple gastric cancer tissue samples and the second Raman spectrum detection data corresponding to the non-gastric cancer tissue samples.

3. The method according to claim 2, characterized in that The step of dividing the Raman spectrum detection data into a model training set and a model verification set comprises: The first Raman spectrum detection data and the second Raman spectrum detection data are respectively divided into a model training set and a model verification set according to a preset ratio, the model training set contains the first Raman spectrum detection data and the second Raman spectrum detection data in a preset ratio, and the model verification set contains the first Raman spectrum detection data and the second Raman spectrum detection data in a preset ratio.

4. The method according to claim 3, characterized in that The step of training the initial model based on the model training set to obtain the trained initial model includes: The Raman spectrum detection data in the model training set and the target gastric disease symptom corresponding to each piece of Raman spectrum detection data are input into the initial model, and the initial model is trained to obtain a trained initial model.

5. The method according to claim 4, characterized in that The step of evaluating the model validation set based on the trained initial model to obtain an evaluation result includes: Inputting the model verification set into the trained initial model so that the trained initial model evaluates the model verification set to obtain an evaluation result; Determine the validity of the evaluation results.

6. The method according to claim 5, characterized in that When the validity of the evaluation result meets the preset condition, determining that the initial model training is completed, and using the trained initial model as a gastric tissue pathology risk evaluation model, including: If the validity of the evaluation result does not meet the preset condition, re-executing the step of training the initial model based on the model training set to obtain the trained initial model; If the validity of the evaluation result meets the preset conditions, it is determined that the initial model training is completed, and the trained initial model is used as the gastric tissue pathology risk evaluation model.

7. A gastric tissue pathology risk assessment device, characterized in that: include: An acquisition module, used for acquiring target Raman spectrum detection data of a gastric tissue sample to be evaluated of a current patient; An evaluation module, used for inputting the target Raman spectrum detection data into a gastric tissue pathology risk evaluation model to obtain an output result of the gastric tissue pathology risk evaluation model; A determination module, configured to determine a gastric tissue pathology risk assessment result of the current patient based on the output result; A model training module is used to obtain Raman spectral detection data corresponding to multiple gastric tissue samples, wherein each piece of Raman spectral detection data corresponds to a target gastric disease symptom; divide the Raman spectral detection data into a model training set and a model verification set; train the initial model based on the model training set to obtain a trained initial model; evaluate the model verification set based on the trained initial model to obtain an evaluation result; when the validity of the evaluation result meets a preset condition, determine that the initial model training is completed, and use the trained initial model as a gastric tissue pathology risk assessment model.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the gastric tissue pathology risk assessment method according to any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the gastric tissue pathology risk assessment method according to any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the gastric tissue pathology risk assessment method according to any one of claims 1 to 6 is implemented.

Images