Digital five-step disease management model system developed based on AIGC technology

Abstract

The present invention discloses a digital five-step disease management model system developed based on AIGC technology, which relates to the field of medical information, an AIGC auxiliary module, which builds an AIGC model architecture; a disease education module, which generates a disease education program corresponding to a disease type; a disease inspection module, which generates a disease inspection video corresponding to a disease type; a disease assessment module, which generates a disease assessment report for a patient; a disease intervention module, which generates a disease intervention program; and a disease follow-up module, which generates a psychological counseling video. By setting an AIGC auxiliary module, a disease education module, a disease inspection module, a disease assessment module, a disease intervention module, and a disease follow-up module, the guidance efficiency is greatly improved, and a corresponding guidance program or video is made using the five-step method of education, inspection, assessment, intervention, and follow-up.

Description

Technical Field

[0001] The present invention relates to the field of medical information, and in particular to a digital five-step disease management model system developed based on AIGC technology. Background Art

[0002] Disease management is another major strategy of health management, which has a long history of development. Disease management is a system that coordinates medical care interventions and communication with patients, and it emphasizes the importance of patient self-care. Disease management supports the doctor-patient relationship and health care plans, emphasizes the use of evidence-based medicine and strategies to enhance personal capabilities to prevent the deterioration of the disease, and evaluates clinical, humanistic and economic effects based on the continuous improvement of individual or overall health.

[0003] Existing disease management must be guided by evidence-based medicine and coordinated between doctors and patients. When the number of patients is large and there is a shortage of doctors, the waiting time for patients will increase and the timeliness of disease management cannot be met. In addition, the existing disease management methods are relatively simple and the guidance effect is insufficient. Summary of the invention

[0004] In order to solve the above technical problems, a digital five-step disease management model system developed based on AIGC technology is provided. This technical solution solves the problem proposed in the above background technology that the existing disease management must be guided by evidence-based medicine and coordinated by doctors and patients. When the number of patients is large and the doctors are insufficient, the waiting time for patients will increase, and the timeliness of disease management cannot be met. In addition, the existing disease management methods are relatively single and the guidance effect is insufficient.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0006] The digital five-step disease management model system developed based on AIGC technology includes:

[0007] An AIGC auxiliary module, wherein the AIGC auxiliary module builds an AIGC model architecture, and the AIGC model architecture is divided into a text generation architecture, an image generation architecture, a speech generation architecture, and a video generation architecture;

[0008] Disease education module, the disease education module uses the AIGC model architecture to train and generate a disease education model, the disease education module obtains characteristic diseases, uses the disease education model to generate a disease education plan corresponding to the disease type, and sends the disease education plan to the patient, the disease education plan includes the basic symptoms, complications and consequences of the characteristic disease;

[0009] Disease inspection module, the disease inspection module uses the AIGC model architecture to train and generate a disease inspection model, the disease inspection module obtains characteristic diseases, uses the disease inspection model to generate a disease inspection video corresponding to the disease type, examines the patient according to the disease inspection video, and obtains the disease inspection result;

[0010] A disease assessment module, which uses the AIGC model architecture to train and generate a disease assessment model. The disease assessment module obtains disease examination results and uses the disease assessment model to generate a disease assessment report for the patient;

[0011] A disease intervention module, wherein the disease intervention module uses the AIGC model architecture to train and generate a disease intervention model. The disease intervention module obtains a disease assessment report and uses the disease intervention model to generate a disease intervention plan. The disease intervention plan includes drug treatment, diet control, and physical exercise recommendation items.

[0012] The disease follow-up module uses the AIGC model architecture to train and generate a disease follow-up model. The disease follow-up module obtains a disease assessment report, uses the disease follow-up model to generate a psychological counseling video, and sends the psychological counseling video to the patient.

[0013] Preferably, the AIGC auxiliary module builds the AIGC model architecture including the following steps:

[0014] When building the text generation architecture, design the text initial neural network model, build the model training platform, and build the model feedback mechanism. The model feedback mechanism adds neurons in the hidden layer and output layer of the text initial neural network model.

[0015] When building the image generation architecture, based on the text generation architecture, design the initial neural network model for converting text into images, build a model training platform, and build a model feedback mechanism. The model feedback mechanism adds neurons to the hidden layer and output layer of the initial neural network model for converting text into images.

[0016] When building the speech generation architecture, we build a platform for text recognition into speech based on the text generation architecture.

[0017] When building the video generation architecture, build a frame-by-frame image merging platform based on the image generation architecture.

[0018] Preferably, the disease education module uses the AIGC model architecture, and training and generating the disease education model includes the following steps:

[0019] Call the text generation framework;

[0020] The model training platform obtains at least one disease sample data and at least one education sample program, and the disease sample data and the education sample program correspond one to one;

[0021] Use disease sample data to train the text initial neural network model to obtain educational text training results;

[0022] Determine whether the overlap between the educational text training results and the educational sample scheme corresponding to the disease sample data is greater than the preset value. If not, use the text initial neural network model as the disease education model. If so, use the model feedback mechanism to repair the text initial neural network model and replace the original text initial neural network model, and repeat the previous step.

[0023] Preferably, the disease detection module uses the AIGC model architecture, and training and generating the disease detection model includes the following steps:

[0024] Calling text generation architecture, image generation architecture, speech generation architecture and video generation architecture;

[0025] The model training platform obtains at least one disease sample data and at least one disease examination sample video, and the disease sample data and the disease examination sample video correspond one to one;

[0026] Use disease sample data to train the text initial neural network model to obtain disease examination text training results;

[0027] Using the disease examination text training results, an initial neural network model for converting text into images is trained to obtain at least one disease examination image;

[0028] The text recognition to speech platform converts the disease examination text training result into disease examination speech, and the image frame-by-frame merging platform merges at least one disease examination image and the disease examination speech to obtain a disease examination training video;

[0029] Determine whether the overlap between the disease inspection training video and the disease inspection sample video corresponding to the disease sample data is greater than a preset value. If not, use the text initial neural network model and the initial neural network model for converting text into image as the disease inspection model. If so, use the model feedback mechanism to repair the text initial neural network model and the initial neural network model for converting text into image and replace the original text initial neural network model and the initial neural network model for converting text into image, repeat the third step, and continue with subsequent steps in sequence.

[0030] Preferably, the disease assessment module uses the AIGC model architecture, and training and generating the disease assessment model includes the following steps:

[0031] Call the text generation framework;

[0032] The model training platform obtains at least one disease sample data and at least one disease assessment sample scheme, and the disease sample data and the disease assessment sample scheme correspond one to one;

[0033] Use disease sample data to train the text initial neural network model to obtain disease assessment text training results;

[0034] Determine whether the overlap between the disease assessment text training results and the disease assessment sample solutions corresponding to the disease sample data is greater than a preset value. If not, use the text initial neural network model as the disease assessment model. If so, use the model feedback mechanism to repair the text initial neural network model and replace the original text initial neural network model, and repeat the previous step.

[0035] Preferably, the disease intervention module uses the AIGC model architecture, and training and generating the disease intervention model includes the following steps:

[0036] Call the text generation framework;

[0037] The model training platform obtains at least one disease sample data and at least one disease intervention sample plan, and the disease sample data and the disease intervention sample plan correspond one to one;

[0038] Use disease sample data to train the text initial neural network model to obtain disease intervention text training results;

[0039] Determine whether the overlap between the disease intervention text training results and the disease intervention sample plan corresponding to the disease sample data is greater than the preset value. If not, the text initial neural network model is used as the disease intervention model. If so, the model feedback mechanism is used to repair the text initial neural network model and replace the original text initial neural network model, and repeat the previous step.

[0040] Preferably, the disease follow-up module uses the AIGC model architecture, and training and generating the disease follow-up model includes the following steps:

[0041] Calling text generation architecture, image generation architecture, speech generation architecture and video generation architecture;

[0042] The model training platform obtains at least one disease sample data and at least one disease follow-up sample video, and the disease sample data and the disease follow-up sample video correspond one to one;

[0043] Use disease sample data to train the text initial neural network model to obtain disease follow-up text training results;

[0044] Using the disease follow-up text training results to train an initial neural network model for converting text into images, to obtain at least one disease follow-up image;

[0045] The text recognition to speech platform converts the disease follow-up text training results into disease follow-up speech, and the image frame-by-frame merging platform merges at least one disease follow-up image and the disease follow-up speech to obtain a disease follow-up training video;

[0046] Determine whether the overlap between the disease follow-up training video and the disease follow-up sample video corresponding to the disease sample data is greater than a preset value. If not, use the text initial neural network model and the initial neural network model for converting text into image as the disease follow-up model. If so, use the model feedback mechanism to repair the text initial neural network model and the initial neural network model for converting text into image and replace the original text initial neural network model and the initial neural network model for converting text into image, repeat the third step, and continue with subsequent steps in sequence.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] By setting up AIGC auxiliary modules, disease education modules, disease examination modules, disease assessment modules, disease intervention modules and disease follow-up modules, and using AIGC technology to automatically generate guidance plans in disease management, the guidance efficiency can be greatly improved, and the waiting time for patients can be reduced. It also fully considers the complexity of disease management, and uses the five-step method of education, examination, assessment, intervention and follow-up to comprehensively manage the disease. It fully considers the various situations that patients need to deal with in disease management, and makes corresponding guidance plans or videos so that patients can better cope with the disease. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the structure of the digital five-step disease management model system of the present invention;

[0050] Figure 2 A schematic diagram of the AIGC model architecture process for the AIGC auxiliary module of the present invention;

[0051] Figure 3 A flowchart of a disease education model generated by training using the AIGC model architecture for the disease education module of the present invention;

[0052] Figure 4 A schematic diagram of the process of training and generating a disease detection model using the AIGC model architecture for the disease detection module of the present invention;

[0053] Figure 5 A schematic diagram of a process flow of training and generating a disease assessment model using the AIGC model architecture for the disease assessment module of the present invention;

[0054] Figure 6 The disease intervention module of the present invention uses the AIGC model architecture to train and generate a disease intervention model flow chart;

[0055] Figure 7 The disease follow-up module of the present invention uses the AIGC model architecture to train and generate a disease follow-up model flow chart. DETAILED DESCRIPTION

[0056] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0057] Reference Figure 1 As shown in the figure, the digital five-step disease management model system developed based on AIGC technology includes:

[0058] An AIGC auxiliary module, wherein the AIGC auxiliary module builds an AIGC model architecture, and the AIGC model architecture is divided into a text generation architecture, an image generation architecture, a speech generation architecture, and a video generation architecture;

[0059] Disease education module, the disease education module uses the AIGC model architecture to train and generate a disease education model, the disease education module obtains characteristic diseases, uses the disease education model to generate a disease education plan corresponding to the disease type, and sends the disease education plan to the patient, the disease education plan includes the basic symptoms, complications and consequences of the characteristic disease;

[0060] Disease inspection module, the disease inspection module uses the AIGC model architecture to train and generate a disease inspection model, the disease inspection module obtains characteristic diseases, uses the disease inspection model to generate a disease inspection video corresponding to the disease type, examines the patient according to the disease inspection video, and obtains the disease inspection result;

[0061] The disease examination results are the severity of the characteristic disease;

[0062] When using the disease inspection model, the text in the disease inspection model is used to initialize the neural network model to obtain the actual text of the disease inspection;

[0063] Using actual disease examination text input into the initial neural network model that converts text into images in the disease examination model, obtaining at least one actual disease examination image;

[0064] The text recognition to speech platform converts the actual text of the disease examination into the actual speech of the disease examination, and the frame-by-frame image merging platform merges at least one actual image of the disease examination and the actual speech of the disease examination to obtain a disease examination video;

[0065] A disease assessment module, which uses the AIGC model architecture to train and generate a disease assessment model. The disease assessment module obtains disease examination results and uses the disease assessment model to generate a disease assessment report for the patient;

[0066] A disease intervention module, wherein the disease intervention module uses the AIGC model architecture to train and generate a disease intervention model. The disease intervention module obtains a disease assessment report and uses the disease intervention model to generate a disease intervention plan. The disease intervention plan includes drug treatment, diet control, and physical exercise recommendation items.

[0067] A disease follow-up module, wherein the disease follow-up module uses the AIGC model architecture to train and generate a disease follow-up model. The disease follow-up module obtains a disease assessment report, uses the disease follow-up model to generate a psychological counseling video, and sends the psychological counseling video to the patient;

[0068] When using the disease follow-up model, the text initial neural network model in the disease follow-up model is used to obtain the actual text of the disease follow-up;

[0069] Using actual disease follow-up text input into an initial neural network model that converts text into images in a disease follow-up model to obtain at least one actual disease follow-up image;

[0070] The text recognition to speech platform converts the actual text of the disease follow-up into the actual speech of the disease follow-up, and the frame-by-frame image merging platform merges at least one actual image of the disease follow-up and the actual speech of the disease follow-up to obtain a psychological counseling video.

[0071] Reference Figure 2 As shown in the figure, the AIGC auxiliary module builds the AIGC model architecture including the following steps:

[0072] When building the text generation architecture, design the text initial neural network model, build the model training platform, and build the model feedback mechanism. The model feedback mechanism adds neurons in the hidden layer and output layer of the text initial neural network model.

[0073] When building the image generation architecture, based on the text generation architecture, design the initial neural network model for converting text into images, build a model training platform, and build a model feedback mechanism. The model feedback mechanism adds neurons to the hidden layer and output layer of the initial neural network model for converting text into images.

[0074] When building the speech generation architecture, we build a platform for text recognition into speech based on the text generation architecture.

[0075] When building the video generation architecture, build an image frame-by-frame merging platform based on the image generation architecture;

[0076] The text generation architecture, image generation architecture, speech generation architecture, and video generation architecture are used in conjunction with each other to generate corresponding text or video to guide patients.

[0077] Reference Figure 3 As shown, the disease education module uses the AIGC model architecture, and training and generating the disease education model includes the following steps:

[0078] Call the text generation framework;

[0079] The model training platform obtains at least one disease sample data and at least one education sample program, and the disease sample data and the education sample program correspond one to one;

[0080] Use disease sample data to train the text initial neural network model to obtain educational text training results;

[0081] Determine whether the overlap between the educational text training results and the educational sample scheme corresponding to the disease sample data is greater than a preset value. If not, use the text initial neural network model as the disease education model. If so, use the model feedback mechanism to repair the text initial neural network model and replace the original text initial neural network model, and repeat the previous step.

[0082] The role of the disease education model is to generate disease education programs corresponding to the types of diseases, so that patients have a basic understanding and knowledge of their own diseases, the causes, symptoms and possible complications of the disease, and enhance patients' compliance with the treatment plan.

[0083] Reference Figure 4 As shown, the disease detection module uses the AIGC model architecture, and training and generating the disease detection model includes the following steps:

[0084] Calling text generation architecture, image generation architecture, speech generation architecture and video generation architecture;

[0085] The model training platform obtains at least one disease sample data and at least one disease examination sample video, and the disease sample data and the disease examination sample video correspond one to one;

[0086] Use disease sample data to train the text initial neural network model to obtain disease examination text training results;

[0087] Using the disease examination text training results, an initial neural network model for converting text into images is trained to obtain at least one disease examination image;

[0088] The text recognition to speech platform converts the disease examination text training result into disease examination speech, and the image frame-by-frame merging platform merges at least one disease examination image and the disease examination speech to obtain a disease examination training video;

[0089] Determine whether the overlap between the disease inspection training video and the disease inspection sample video corresponding to the disease sample data is greater than a preset value. If not, use the text initial neural network model and the initial neural network model of text-to-image conversion as the disease inspection model. If so, use the model feedback mechanism to repair the text initial neural network model and the initial neural network model of text-to-image conversion and replace the original text initial neural network model and the initial neural network model of text-to-image conversion, repeat the third step, and continue with the subsequent steps in sequence.

[0090] The function of the disease examination model is to generate disease examination videos corresponding to the disease types, so that doctors can examine patients based on the videos to prevent omissions in doctor's examinations.

[0091] Reference Figure 5 As shown, the disease assessment module uses the AIGC model architecture, and training and generating the disease assessment model includes the following steps:

[0092] Call the text generation framework;

[0093] The model training platform obtains at least one disease sample data and at least one disease assessment sample scheme, and the disease sample data and the disease assessment sample scheme correspond one to one;

[0094] Use disease sample data to train the text initial neural network model to obtain disease assessment text training results;

[0095] Determine whether the overlap between the disease assessment text training results and the disease assessment sample solutions corresponding to the disease sample data is greater than a preset value. If not, use the text initial neural network model as the disease assessment model. If so, use the model feedback mechanism to repair the text initial neural network model and replace the original text initial neural network model, and repeat the previous step.

[0096] The function of the disease assessment model is to generate a disease assessment report for the patient, so that targeted disease intervention can be carried out based on the disease assessment report to prevent the patient's disease from worsening.

[0097] Reference Figure 6 As shown, the disease intervention module uses the AIGC model architecture, and training and generating the disease intervention model includes the following steps:

[0098] Call the text generation framework;

[0099] The model training platform obtains at least one disease sample data and at least one disease intervention sample plan, and the disease sample data and the disease intervention sample plan correspond one to one;

[0100] Use disease sample data to train the text initial neural network model to obtain disease intervention text training results;

[0101] Determine whether the overlap between the disease intervention text training results and the disease intervention sample plan corresponding to the disease sample data is greater than the preset value. If not, the text initial neural network model is used as the disease intervention model. If so, the model feedback mechanism is used to repair the text initial neural network model and replace the original text initial neural network model, and repeat the previous step.

[0102] Reference Figure 7 As shown, the disease follow-up module uses the AIGC model architecture, and training and generating the disease follow-up model includes the following steps:

[0103] Calling text generation architecture, image generation architecture, speech generation architecture and video generation architecture;

[0104] The model training platform obtains at least one disease sample data and at least one disease follow-up sample video, and the disease sample data and the disease follow-up sample video correspond one to one;

[0105] Use disease sample data to train the text initial neural network model to obtain disease follow-up text training results;

[0106] Using the disease follow-up text training results to train an initial neural network model for converting text into images, to obtain at least one disease follow-up image;

[0107] The text recognition to speech platform converts the disease follow-up text training results into disease follow-up speech, and the image frame-by-frame merging platform merges at least one disease follow-up image and the disease follow-up speech to obtain a disease follow-up training video;

[0108] Determine whether the overlap between the disease follow-up training video and the disease follow-up sample video corresponding to the disease sample data is greater than a preset value. If not, use the text initial neural network model and the initial neural network model of text-to-image conversion as the disease follow-up model. If so, use the model feedback mechanism to repair the text initial neural network model and the initial neural network model of text-to-image conversion and replace the original text initial neural network model and the initial neural network model of text-to-image conversion, repeat the third step, and continue with the subsequent steps in sequence.

[0109] The function of the disease follow-up model is to generate psychological counseling videos. After a patient becomes ill, his or her emotions will become abnormal due to his or her own condition, which will not only affect the condition, but also may cause the patient to refuse to cooperate with the treatment. Therefore, psychological counseling videos are needed to soothe the patient's emotions.

[0110] Furthermore, the present solution also proposes a storage medium on which a computer-readable program is stored. When the computer-readable program is called, it runs the above-mentioned digital five-step disease management model system developed based on AIGC technology.

[0111] It is understandable that the storage medium may be a magnetic medium, such as a floppy disk, a hard disk, or a magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid state drive (SSD).

[0112] To sum up, the advantages of the present invention are: by setting up AIGC auxiliary module, disease education module, disease examination module, disease assessment module, disease intervention module and disease follow-up module, using AIGC technology to automatically generate guidance plans in disease management, can greatly improve guidance efficiency, reduce patient waiting time, and fully consider the complexity of disease management, use education, examination, assessment, intervention and follow-up five-step method to comprehensively manage the disease, fully consider the various situations that patients need to deal with in disease management, and make corresponding guidance plans or videos so that patients can better cope with the disease.

[0113] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A digital five-step disease management model system developed based on AIGC technology, characterized by: include: An AIGC auxiliary module, wherein the AIGC auxiliary module builds an AIGC model architecture, and the AIGC model architecture is divided into a text generation architecture, an image generation architecture, a speech generation architecture, and a video generation architecture; Disease education module, the disease education module uses the AIGC model architecture to train and generate a disease education model, the disease education module obtains characteristic diseases, uses the disease education model to generate a disease education plan corresponding to the disease type, and sends the disease education plan to the patient, the disease education plan includes the basic symptoms, complications and consequences of the characteristic disease; Disease inspection module, the disease inspection module uses the AIGC model architecture to train and generate a disease inspection model, the disease inspection module obtains characteristic diseases, uses the disease inspection model to generate a disease inspection video corresponding to the disease type, examines the patient according to the disease inspection video, and obtains the disease inspection result; A disease assessment module, which uses the AIGC model architecture to train and generate a disease assessment model. The disease assessment module obtains disease examination results and uses the disease assessment model to generate a disease assessment report for the patient; A disease intervention module, wherein the disease intervention module uses the AIGC model architecture to train and generate a disease intervention model. The disease intervention module obtains a disease assessment report and uses the disease intervention model to generate a disease intervention plan. The disease intervention plan includes drug treatment, diet control, and physical exercise recommendation items. A disease follow-up module, wherein the disease follow-up module uses the AIGC model architecture to train and generate a disease follow-up model. The disease follow-up module obtains a disease assessment report, uses the disease follow-up model to generate a psychological counseling video, and sends the psychological counseling video to the patient; The disease assessment module uses the AIGC model architecture, and training and generating the disease assessment model includes the following steps: Call the text generation framework; The model training platform obtains at least one disease sample data and at least one disease assessment sample scheme, and the disease sample data and the disease assessment sample scheme correspond one to one; Use disease sample data to train the text initial neural network model to obtain disease assessment text training results; Determine whether the overlap between the disease assessment text training results and the disease assessment sample solutions corresponding to the disease sample data is greater than a preset value. If not, use the text initial neural network model as the disease assessment model. If so, use the model feedback mechanism to repair the text initial neural network model and replace the original text initial neural network model, and repeat the previous step. The disease intervention module uses the AIGC model architecture, and training and generating the disease intervention model includes the following steps: Call the text generation framework; The model training platform obtains at least one disease sample data and at least one disease intervention sample plan, and the disease sample data and the disease intervention sample plan correspond one to one; Use disease sample data to train the text initial neural network model to obtain disease intervention text training results; Determine whether the overlap between the disease intervention text training results and the disease intervention sample scheme corresponding to the disease sample data is greater than a preset value. If not, use the text initial neural network model as the disease intervention model. If so, use the model feedback mechanism to repair the text initial neural network model and replace the original text initial neural network model, and repeat the previous step. The disease follow-up module uses the AIGC model architecture, and training and generating the disease follow-up model includes the following steps: Calling text generation architecture, image generation architecture, speech generation architecture and video generation architecture; The model training platform obtains at least one disease sample data and at least one disease follow-up sample video, and the disease sample data and the disease follow-up sample video correspond one to one; Use disease sample data to train the text initial neural network model to obtain disease follow-up text training results; Using the disease follow-up text training results to train an initial neural network model for converting text into images, to obtain at least one disease follow-up image; The text recognition to speech platform converts the disease follow-up text training results into disease follow-up speech, and the image frame-by-frame merging platform merges at least one disease follow-up image and the disease follow-up speech to obtain a disease follow-up training video; Determine whether the overlap between the disease follow-up training video and the disease follow-up sample video corresponding to the disease sample data is greater than a preset value. If not, use the text initial neural network model and the initial neural network model for converting text into image as the disease follow-up model. If so, use the model feedback mechanism to repair the text initial neural network model and the initial neural network model for converting text into image and replace the original text initial neural network model and the initial neural network model for converting text into image, repeat the third step, and continue with subsequent steps in sequence.

2. According to claim 1, the digital five-step disease management model system developed based on AIGC technology is characterized in that: The AIGC auxiliary module builds the AIGC model architecture including the following steps: When building the text generation architecture, design the text initial neural network model, build the model training platform, and build the model feedback mechanism. The model feedback mechanism adds neurons in the hidden layer and output layer of the text initial neural network model. When building the image generation architecture, based on the text generation architecture, design the initial neural network model for converting text into images, build a model training platform, and build a model feedback mechanism. The model feedback mechanism adds neurons to the hidden layer and output layer of the initial neural network model for converting text into images. When building the speech generation architecture, we build a platform for text recognition into speech based on the text generation architecture. When building the video generation architecture, build a frame-by-frame image merging platform based on the image generation architecture.

3. The digital five-step disease management model system developed based on AIGC technology according to claim 2 is characterized in that: The disease education module uses the AIGC model architecture, and training and generating the disease education model includes the following steps: Call the text generation framework; The model training platform obtains at least one disease sample data and at least one education sample program, and the disease sample data and the education sample program correspond one to one; Use disease sample data to train the text initial neural network model to obtain educational text training results; Determine whether the overlap between the educational text training results and the educational sample scheme corresponding to the disease sample data is greater than the preset value. If not, use the text initial neural network model as the disease education model. If so, use the model feedback mechanism to repair the text initial neural network model and replace the original text initial neural network model, and repeat the previous step.

4. The digital five-step disease management model system developed based on AIGC technology according to claim 3 is characterized in that: The disease detection module uses the AIGC model architecture, and training and generating the disease detection model includes the following steps: Calling text generation architecture, image generation architecture, speech generation architecture and video generation architecture; The model training platform obtains at least one disease sample data and at least one disease examination sample video, and the disease sample data and the disease examination sample video correspond one to one; Use disease sample data to train the text initial neural network model to obtain disease examination text training results; Using the disease examination text training results, an initial neural network model for converting text into images is trained to obtain at least one disease examination image; The text recognition to speech platform converts the disease examination text training result into disease examination speech, and the image frame-by-frame merging platform merges at least one disease examination image and the disease examination speech to obtain a disease examination training video; Determine whether the overlap between the disease inspection training video and the disease inspection sample video corresponding to the disease sample data is greater than a preset value. If not, use the text initial neural network model and the initial neural network model for converting text into image as the disease inspection model. If so, use the model feedback mechanism to repair the text initial neural network model and the initial neural network model for converting text into image and replace the original text initial neural network model and the initial neural network model for converting text into image, repeat the third step, and continue with subsequent steps in sequence.

Images