Ostomy abnormal state online identification and diagnosis evaluation system and method

Abstract

The invention relates to a stoma abnormal state on-line identification and diagnosis evaluation system and method, a patient photographs a stoma part through a patient end APP, a stoma picture is collected and uploaded to a background server, the background server performs stoma abnormal intelligent evaluation on the stoma picture, and the stoma abnormal state on-line identification and diagnosis evaluation is performed on the stoma abnormal state. Whether the stoma is abnormal or not is judged, emergency propaganda and education measures corresponding to the abnormal stoma are issued, and the patient can conduct emergency treatment on the abnormal stoma in time according to the emergency propaganda and education measures. By identifying and diagnosing the scanned stoma data of the patient, judging whether the stoma data is abnormal or not and judging the type of the abnormal stoma data, and intelligently giving the corresponding stoma processing suggestions, the automatic monitoring of the stoma abnormality at home of the patient is realized, the abnormal condition is propagandized, taught and nursed in time, and the risk of the patient caused by corresponding complications is avoided. And the patient can learn online propaganda and education, so that the patient can timely supplement stoma nursing knowledge, and the risk control degree is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of clinical intelligent recognition and analysis technologies, and particularly to an online recognition and diagnosis evaluation system for stoma abnormal states, a stoma abnormal evaluation method, and an electronic device. Background Art

[0002] A stoma refers to a situation where, due to digestive or urinary system diseases, surgical treatment is required to separate the intestinal tract, and one end of the intestinal tract is led out to the body surface (such as the Figure 1 shown anal stoma) to form an opening. A section of the intestinal tract above the rectum is pulled out, opened outside the body, and turned and sutured and fixed to the abdominal wall to form an artificial anus, which is used to replace the defecation function of the original rectal anus.

[0003] A stoma can achieve intestinal decompression, relieve obstruction, protect the anastomosis or injury of the distal intestinal tract, promote the recovery of intestinal and urinary tract diseases, and even save the patient's life. The postoperative care of a stoma mainly involves the following two aspects:

[0004] One is the precautions for stoma care:

[0005] 1. Observe the blood circulation of the stoma and the recovery of the stoma, such as whether there is stoma contraction, bleeding, etc.

[0006] 2. Observe the changes, color, and properties of the liquid in the stoma bag in a timely manner.

[0007] 3. After suture removal, perform appropriate anal dilation to avoid stoma stenosis.

[0008] 4. Pay attention to the care of the skin around the stoma to avoid intestinal fluid from irritating the skin and causing eczema and other skin diseases.

[0009] The other is stoma complications

[0010] 1. Infection around the stoma: Due to inadequate disinfection treatment at the stoma site, inflammation and infection around the stoma may occur.

[0011] 2. Stoma skin eczema or dermatitis: Due to repeated and continuous irritation of the skin by stoma fluid, local eczema or stubborn dermatitis may occur.

[0012] 3. Other complications also include stoma stenosis, stoma bleeding, stoma retraction, electrolyte imbalance, local ischemia and necrosis, and recurrence of tumors at the stoma site, etc.

[0013] Although patients can receive timely examination and diagnosis of stomas by medical staff in the hospital, when patients are discharged and cared for at home, they cannot screen the stoma status by themselves and cannot identify abnormal or normal stomas. Therefore, ineffective stoma home monitoring will bring different complications to stoma patients, and even more serious life-threatening risks.

[0014] As people's living standards improve, they pay more and more attention to health, and the popularity of colorectal cancer and bladder cancer treatment surgeries is becoming more and more widespread. More and more patients are carrying stomas (including temporary stomas and permanent stomas) after surgery. However, there are still many stoma patients who need knowledge and skills training for stoma maintenance, early identification and treatment of stoma complications, but due to various factors such as inconvenient transportation, information barriers, economic problems, and cultural level, it is difficult to get effective help in a timely manner. Summary of the invention

[0015] In order to solve the above problems, the present application proposes an online recognition and diagnosis evaluation system for stoma abnormality, a stoma abnormality evaluation method and an electronic device.

[0016] On the one hand, this application proposes an online recognition and diagnosis evaluation system for stoma abnormality, including a patient-side APP, a doctor's PC and a backend server, wherein:

[0017] The patient-side APP is used for patients to log in to the backend server and upload their own stoma pictures;

[0018] The backend server is used to perform an intelligent evaluation of stoma abnormality on the stoma image to determine whether the stoma is abnormal:

[0019] If it is abnormal, an abnormal stoma warning notification will be sent to the patient APP, and emergency education measures corresponding to the abnormal stoma will be issued;

[0020] If there is no abnormality, a normal stoma notification is sent to the patient APP;

[0021] The doctor's PC is used for the doctor to log in to the backend server, follow up on the stoma pictures and abnormality assessment results of each patient, and determine whether to issue a medical order. If a medical order is issued, the medical order is forwarded to the corresponding patient-side APP through the backend server;

[0022] The patient-side APP and the doctor's PC are respectively connected to the backend server for communication.

[0023] As an optional implementation scheme of the present application, optionally, the background server includes:

[0024] API interface, used for APP / PC access and data interaction;

[0025] A medical care management system for managing the doctor's PC terminal;

[0026] A patient management system for managing the patient-side APP and binding the stoma pictures and their abnormal assessment results of each patient to the medical record ID of the corresponding patient;

[0027] A MYSQL database for storing the stoma pictures and their abnormal assessment results of each patient, and storing the emergency education measures for different abnormal stomas;

[0028] A stoma education module for saving the mapping relationship between different abnormal stomas and the corresponding emergency education measures;

[0029] A stoma abnormality intelligent assessment module for using a pre-deployed stoma abnormality AI recognition model to identify stoma abnormalities in the stoma pictures of patients and assessing whether a preset abnormal stoma feature image appears in the stoma pictures of patients:

[0030] If so, output the stoma abnormality recognition result and bind the stoma abnormality recognition result to the medical record ID of the corresponding patient by the patient management system;

[0031] Otherwise, output the stoma normal recognition result and bind the stoma normal recognition result to the medical record ID of the corresponding patient by the patient management system.

[0032] As an optional implementation of this application, optionally, the method for generating the stoma abnormality AI recognition model includes:

[0033] Collect a number of stoma abnormality pictures;

[0034] Perform feature engineering on each stoma abnormality picture, extract the abnormal stoma feature images on each stoma abnormality picture, and form an abnormal stoma feature image set;

[0035] Divide the abnormal stoma feature image set into a training set and a validation set according to a preset ratio;

[0036] Import the training set into a preset convolutional neural network model for deep learning training to generate the stoma abnormality AI recognition model;

[0037] Use the validation set to verify whether the stoma abnormality AI recognition model meets the standard;

[0038] If the verification is qualified, deploy the stoma abnormality AI recognition model on the background server;

[0039] If the verification is unqualified, re-perform feature engineering and re-train the model.

[0040] As an alternative implementation of the present application, optionally, after verifying whether the stoma abnormality AI recognition model meets the standard by using the verification set, the following steps are further included:

[0041] Prepare several normal stoma pictures;

[0042] Input the normal stoma pictures into the stoma abnormality AI recognition model to verify whether the stoma abnormality AI recognition model generates an image feature recognition response to the normal stoma pictures:

[0043] If a response is generated, re-collect the abnormal stoma pictures and retrain the model;

[0044] Otherwise, proceed to the next step.

[0045] As an alternative implementation of the present application, optionally, the patient management system is further configured to:

[0046] Retrieve the emergency education measures for different abnormal stomas from the MYSQL database at a preset frequency and send them to the patient-side APP to notify the patient to conduct stoma care education learning regularly.

[0047] As an alternative implementation of the present application, optionally, the patient-side APP is further configured to:

[0048] Collect the learning logs of the patient's stoma care education learning and upload them to the background server, which is then forwarded by the background server to the patient management system;

[0049] The patient management system is further configured to:

[0050] Bind the patient's learning logs to the corresponding patient's visit ID.

[0051] As an alternative implementation of the present application, optionally, the patient-side APP is further configured to:

[0052] The patient shares daily care information with the background server, which is then forwarded by the background server to the patient management system;

[0053] The patient management system is further configured to:

[0054] Write the patient's daily care information into the patient management system and bind it to the corresponding patient's visit ID.

[0055] As an alternative implementation of the present application, optionally, the doctor's PC terminal is further configured to:

[0056] Log in to the background server, enter the patient management system, view the stoma pictures and their abnormal assessment results of each patient, as well as learning logs and the daily care information; and,

[0057] Generate corresponding medical orders and write them into the patient management system, and the patient management system binds the medical orders under the corresponding patient's visit ID.

[0058] On the other hand, this application proposes a method for stoma abnormal assessment, which is implemented based on the online recognition and diagnosis evaluation system for stoma abnormal status, and includes the following steps:

[0059] The patient takes a photo of the stoma site through the patient-side APP, collects the stoma picture and uploads it to the background server;

[0060] The background server conducts intelligent assessment of stoma abnormalities on the stoma picture to determine whether the stoma is abnormal:

[0061] If it is abnormal, an early warning notice of stoma abnormality is sent to the patient-side APP, and at the same time, emergency education measures for the corresponding abnormal stoma are issued;

[0062] If it is not abnormal, a notice of normal stoma is sent to the patient-side APP;

[0063] The doctor logs in to the background server through the doctor's PC terminal, follows up the stoma pictures and their abnormal assessment results of each patient, determines whether to issue a medical order, and if a medical order is issued, forwards the medical order to the corresponding patient-side APP through the background server.

[0064] On the other hand, this application also proposes an electronic device, including:

[0065] A processor;

[0066] A memory for storing executable instructions of the processor;

[0067] Wherein, the processor is configured to implement the stoma abnormal assessment method when executing the executable instructions.

[0068] The technical effects of the present invention:

[0069] Based on the implementation of the present invention, the patient takes pictures of the stoma site through the patient-side APP, collects stoma pictures and uploads them to the background server. The background server conducts intelligent evaluation of stoma abnormalities on the stoma pictures, determines whether the stoma is abnormal and issues corresponding emergency education measures for the abnormal stoma, enabling the patient to promptly take emergency treatment for the abnormal stoma according to the emergency education measures. By identifying and diagnosing the stoma data scanned by the patient, determining whether it is abnormal and the type of abnormality, and intelligently giving corresponding stoma treatment suggestions, automatic monitoring of stoma abnormalities at home for patients is realized, and emergency education and nursing for abnormal situations are carried out in a timely manner to avoid causing corresponding complications and bringing risks to the patient.

[0070] The present invention can also enable patients to conduct online education and learning on stoma abnormalities and share nursing experiences. Through a comprehensive application system integrating knowledge learning + online diagnosis + medical services + social platforms, the purpose of enabling stoma patients and their families to learn stoma care knowledge, identify stoma abnormalities, receive timely treatment suggestions, communicate with fellow patients to reduce the sense of stigma, and meet the knowledge, skills and spiritual needs of stoma patients is achieved.

[0071] According to the following detailed description of exemplary embodiments with reference to the accompanying drawings, other features and aspects of the present disclosure will become clear. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] The accompanying drawings, which are included in and constitute a part of this specification, illustrate exemplary embodiments, features and aspects of the present disclosure together with the specification and are used to explain the principles of the present disclosure.

[0073] Figure 1 Shown is a schematic diagram of a stoma (artificial anus);

[0074] Figure 2 Shown is a schematic diagram of the composition of the application system of the present invention;

[0075] Figure 3 Shown is a schematic diagram of the training of the stoma abnormality AI recognition model of the present invention;

[0076] Figure 4 Shown is a schematic diagram of the comparison between abnormal and normal stoma images of the present invention;

[0077] Figure 5 Shown is a schematic diagram of the application of the electronic device of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0078] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the drawings denote elements having the same or similar functions. Although various aspects of the embodiments are shown in the drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0079] As used herein, the term "exemplary" means "serving as an example, embodiment, or illustration". Any embodiment described as "exemplary" herein need not be construed as superior or better than other embodiments.

[0080] In addition, for a better illustration of the present disclosure, numerous specific details are given in the following detailed description. Those skilled in the art should understand that the present disclosure can be implemented without some specific details. In some instances, well-known means, elements, and circuits have not been described in detail so as to highlight the gist of the present disclosure.

[0081] Embodiment 1

[0082] In this embodiment, the patient-side APP (mobile terminal), the doctor's PC terminal, and the background server are respectively the interaction devices for patients / their families, doctors, and administrators. The patient-side APP, that is, the mobile terminal, can be an APP installed on a smart phone, and accesses the background server through the API interface of the APP to achieve data interaction; the PC terminal can also allow doctors to access the background, realize the interaction with the background, and manage patients, etc.

[0083] Specifically, it can be understood in combination with the existing hospital interaction system.

[0084] As Figure 2 shown, on the one hand, the present application proposes an online recognition and diagnostic evaluation system for stoma abnormal states, including a patient-side APP, a doctor's PC terminal, and a background server, wherein:

[0085] The patient-side APP is used for patients to log in to the background server and upload their own stoma pictures;

[0086] The background server is used for performing intelligent evaluation of stoma abnormalities on the stoma pictures to determine whether the stoma is abnormal:

[0087] If it is abnormal, a stoma abnormality warning notice is sent to the patient-side APP, and at the same time, emergency education measures corresponding to the abnormal stoma are issued;

[0088] If it is not abnormal, a stoma normal notice is sent to the patient-side APP;

[0089] The doctor's PC terminal is used for doctors to log in to the background server, follow up the stoma pictures and their abnormal evaluation results of each patient, determine whether to issue medical orders, and if medical orders are issued, forward the medical orders to the corresponding patient-side APP through the background server;

[0090] The patient-side APP and the doctor's PC terminal are respectively communicatively connected to the background server.

[0091] The main idea of the present invention is that the mobile terminal scans the stoma and uploads the stoma picture - the background performs AI recognition on the stoma, identifies the abnormal stoma status - outputs treatment opinions:

[0092] It can be handled at home, and treatment suggestions are given (emergency education measures are issued, and the patient takes self-care).

[0093] It cannot be handled at home, and outpatient services or online diagnosis and treatment services of nearby community hospitals are given (the administrator judges whether to issue the nursing opinion of "cannot be handled at home" to the patient according to the abnormal results of the patient and issues it to the patient's APP).

[0094] The patient can also call the camera service through their patient APP to take pictures of their stoma regularly, collect stoma pictures and upload them to the background, and the background performs intelligent recognition on the stoma pictures for online abnormal assessment. There is a stoma abnormality AI recognition model deployed on the background server that can identify whether the stoma is abnormal, and can perform feature recognition and analysis on abnormal stoma pictures to judge whether various abnormal stoma feature images learned by the model appear on the stoma pictures. If abnormal feature images appear, it is determined that there is an abnormal stoma situation on the current stoma picture. Therefore, the background needs to notify the patient of the stoma abnormality and issue a warning notice to them. Specifically, the background server sends a stoma abnormality warning notice to the mobile terminal of the corresponding patient, and at the same time issues corresponding emergency education measures for the patient to perform nursing at home.

[0095] For the patients who receive the abnormal warning notice, the server can also synchronously push the abnormal situation to the PC side of the doctor in charge of the patient, so that the doctor can judge whether the patient can be handled at home. If the doctor believes that it can be handled at home, the corresponding emergency education measures are issued by the background for the patient to take self-care. If the doctor believes that the current abnormal situation is too serious to be handled at home, the doctor notifies the background server through the PC side, and then the background server issues the nursing opinion of "cannot be handled at home" to the patient.

[0096] Regarding the corresponding emergency training measures issued by the system according to the recognition situation of abnormal stomas, the corresponding emergency education measures for different abnormal stomas can be pre-stored on the background server in advance. After the model identifies the feature images of the patient's abnormal stoma, the corresponding emergency education measures can be retrieved and issued according to the type of abnormal stoma. For example, in the case of stoma abnormal necrosis, the emergency education measure at this time is "remove the factors that aggravate ischemia, evaluate the viability and use a transparent stoma bag to reconstruct the stoma", and the emergency education measure for stoma necrosis is issued to the patient for nursing. However, in this case, the doctor will recommend that the patient with stoma necrosis go to the outpatient service of a nearby hospital or online diagnosis and treatment service for diagnosis and treatment.

[0097] For different types of abnormal stoma conditions, corresponding emergency education measures can be set and configured according to the types of stoma abnormalities available clinically, and stored in the MYSQL database on the background server. After the subsequent model identifies the type of stoma abnormality of the current patient, it can automatically match the corresponding emergency selection measures according to the identified type, improving the efficiency of emergency care.

[0098] On the background server, general knowledge points (education data, included in the corresponding emergency education measures) about stoma maintenance, complications, and abnormal stoma status can be imported. Patients can learn through mobile videos about stoma pouch replacement techniques and stoma skin care. This program can also open a social platform for stoma patients to register and communicate with fellow patients about their conditions.

[0099] As an optional implementation of this application, optionally, the background server includes:

[0100] API interface for APP / PC access to achieve data interaction;

[0101] Medical staff management system for managing the doctor's PC terminal;

[0102] Patient management system for managing the patient-side APP, and binding the stoma pictures and their abnormal assessment results of each patient to the corresponding patient's visit ID;

[0103] MYSQL database for storing the stoma pictures and their abnormal assessment results of each patient, and storing the emergency education measures for different abnormal stomas;

[0104] Stoma education module for saving the mapping relationship between different abnormal stomas and the corresponding emergency education measures;

[0105] Stoma abnormality intelligent assessment module for using a pre-deployed stoma abnormality AI recognition model to identify stoma abnormalities in the patient's stoma pictures, and assessing whether a preset abnormal stoma feature image appears in the patient's stoma pictures:

[0106] If it exists, output the stoma abnormality recognition result and have the patient management system bind the stoma abnormality recognition result to the corresponding patient's visit ID;

[0107] Otherwise, output the stoma normal recognition result and have the patient management system bind the stoma normal recognition result to the corresponding patient's visit ID.

[0108] The present invention uses an API interface to achieve communication between patients and doctors. When the backend server communicates with each terminal through the API (Application Programming Interface) interface, it is actually implementing a standardized, cross-platform, request- and response-based communication mechanism. The following are some key points regarding the communication between the backend server and the terminal through the API interface:

[0109] 1. Request and Response:

[0110] The terminal (such as a mobile application, web application, desktop application, etc.) requests data or performs operations by sending HTTP or HTTPS requests to the API interface of the server.

[0111] After receiving the request, the server processes the request and generates a corresponding HTTP response to return to the terminal.

[0112] 2. Standardization:

[0113] API interfaces usually follow one or more standards, such as REST (Representational State Transfer) or SOAP (Simple Object Access Protocol). These standards define how to construct requests and responses, how to transfer data, etc.

[0114] Using standard APIs makes it easier for different terminals to communicate with the server, and also facilitates the maintenance and expansion of the server.

[0115] 3. Cross-platform:

[0116] API interfaces are cross-platform, which means that regardless of what operating system or programming language the terminal uses, as long as it supports HTTP requests, it can communicate with the server.

[0117] This provides great flexibility for developers because they can choose the terminal technology stack that best suits their needs.

[0118] 4. Security:

[0119] The security of API interfaces is very important because they may expose sensitive data or perform sensitive operations.

[0120] Generally, API interfaces use HTTPS to encrypt the transmitted data to prevent the data from being stolen or tampered with during transmission.

[0121] In addition, API interfaces may also use authentication and authorization mechanisms (such as OAuth, JWT, etc.) to verify the identity and permissions of the requester.

[0122] 5. Documentation and Testing:

[0123] A good API interface should have detailed documentation explaining how to use the interface, request parameters, response format, etc.

[0124] Additionally, a testing or sandbox environment should be provided so that developers can test their code without interfering with the production environment.

[0125] 6. Version Control:

[0126] Over time, API interfaces may change or add new features. In order to maintain compatibility with old terminals, version control is usually used to manage different versions of API interfaces.

[0127] The endpoint can select the API interface version to use by specifying the API version in the request.

[0128] 7. Error handling and logging:

[0129] When the request sent by the terminal is invalid or the server cannot process it, the API interface should return an appropriate error code and error message so that the terminal can handle the error accordingly.

[0130] In addition, the server should also record access logs and error logs for all API interfaces so that troubleshooting and auditing can be performed when problems arise.

[0131] 8.Performance optimization:

[0132] For highly concurrent API interfaces, the server needs to be performance optimized to ensure that response time and throughput meet requirements.

[0133] This may include using techniques such as caching, load balancing, asynchronous processing, etc. to reduce server load and improve response time.

[0134] The backend server communicates with each terminal through the API interface in an efficient, flexible and secure way, which enables different terminals to easily interact with the server and obtain the required data or services.

[0135] Because patients need to take pictures of their stomas through mobile devices and upload them to the backend, which will then intelligently identify stoma abnormalities. Therefore, the backend and mobile devices can communicate through API. Doctors need to log in to the backend to view the patient's stoma abnormalities and the results of the backend's intelligent identification and judgment of their stomas, and give corresponding medical advice. Therefore, doctors need to communicate data with the backend through the PC, and also use API data communication to allow doctors and patients to interact with the backend server through the API interface.

[0136] The medical care management system is mainly a system for the background server to manage doctors. When doctors need to retrieve and view various information of patients, they need to send corresponding requests to the patient management system or other system modules through the medical care management system. The background server processes and responds to the requests, enabling doctors to complete various data processing and interactions.

[0137] The patient management system mainly manages each stoma surgery patient, including recording the patient's stoma information, surgical information, etc. through case files. At the same time, it can record the stoma abnormality identification situation of patients and can also manage the stoma education and learning of patients. The patient management system can regularly or irregularly send corresponding stoma care education materials to the mobile terminals where patients are located, notifying patients to conduct stoma emergency care learning regularly, thereby improving the patients' risk control of stomas. During the daily stoma care process, because the patient management system can record the stoma abnormality situations of each patient, it can accurately match the corresponding stoma education materials according to the stoma abnormality types and push the corresponding stoma education materials to the mobile terminals where patients are located, enabling patients to accurately learn and precisely care for their own stoma abnormality situations.

[0138] During the stoma care process, patients can also report and record their daily stoma care situations and experiences, etc. through the mobile terminal (patient-side APP) to the patient management system. The patient management system can manage the communication IDs of different patients and can open access channels to other patients for each patient. Therefore, patient a can send a communication request to any other stoma patient through the patient management system for stoma care sharing. Specifically, a corresponding message middleware can be set on the background server to convey the communication information between the two patients. The message middleware, such as kafka or other message queue mechanisms, can refer to the existing background message distribution management mechanism, which is not limited in this embodiment.

[0139] The database is mainly used for data storage, including storing the stoma pictures collected from patients and the results of abnormal evaluation output for the stoma pictures. At the same time, it can also store the emergency education measures for different abnormal stoma types.

[0140] There is a pre-deployed stoma abnormality AI recognition model in the stoma abnormality intelligent evaluation module, which is generated by training and learning with several different types of abnormal stoma pictures. A deep learning model, such as a convolutional neural network model, is used to train and learn the picture features in different types of abnormal stoma pictures to generate the stoma abnormality AI recognition model, so as to intelligently recognize and evaluate the patient's stoma pictures, and judge whether abnormal stoma features appear on the stoma pictures: if so, it is considered that the current patient has a stoma abnormality; otherwise, the recognition result of a normal stoma is output. The background can register the recognition results output by the evaluation module, write them into the patient management system, and bind them under the corresponding patient visit ID, which can quickly identify and evaluate the patient's stoma abnormality online, avoid the recognition errors and defects caused by the patient's inability to accurately identify the stoma abnormality, and save the patient's time to go to the hospital for treatment.

[0141] In addition, the stoma education module can import various types of stoma bag replacement videos or graphic tutorials to facilitate patients' access to learning.

[0142] The patient management system includes functions such as surgical methods, postoperative duration, stoma types, and patient classification in the attached drawings. An AI scan stoma recognition section is added, which can scan the stoma from various angles to judge whether the stoma is in a normal state, whether there are situations such as collapse, prolapse, and edema, and give suggestions based on the recognition results. Whether it can be handled at home and how to handle it. If the situation is more serious, the patient can be advised to go to the hospital for treatment, and nearby hospitals with medical treatment qualifications can be pushed, or direct push of services such as registration and navigation to the hospital, or push of relevant experts' online consultations, online video and picture consultations.

[0143] The medical staff management interacts with the patient management system. Patients who are good at sharing and optimistic can share their experience of maintaining the stoma on the platform, share the little skills of carrying the stoma to travel in daily life, share the inconveniences brought by the stoma and the treatment measures, give good encouragement to new stoma patients, and provide a communication platform for stoma patients to feel a sense of belonging.

[0144] Such as Figure 3 As shown, as an optional implementation solution of the present application, optionally, the method for generating the stoma abnormality AI recognition model includes:

[0145] Collect several stoma abnormality pictures;

[0146] Perform feature engineering on each of the stoma abnormality pictures, extract the abnormal stoma feature images on each of the stoma abnormality pictures, and form an abnormal stoma feature image set;

[0147] Divide the abnormal stoma feature image set into a training set and a validation set according to a preset ratio;

[0148] Import the training set into a preset convolutional neural network model for deep learning training to generate the stoma anomaly AI recognition model;

[0149] Use the validation set to verify whether the stoma anomaly AI recognition model meets the standards;

[0150] If the verification is successful, deploy the stoma anomaly AI recognition model on the background server;

[0151] If the verification fails, re-perform feature engineering and re-train the model.

[0152] The steps of using a convolutional neural network (CNN) to generate a stoma anomaly recognition model can be summarized as follows:

[0153] 1. Data collection and preprocessing:

[0154] Collect a dataset containing stoma anomaly status (several stoma anomaly pictures).

[0155] Preprocess the data, which usually includes data cleaning (removing noise and outliers), feature extraction (converting the original data into more representative features), and normalization (mapping the data to a unified range to improve the model training effect).

[0156] 2. Construct a convolutional neural network model:

[0157] Design the structure of the convolutional neural network, which usually includes an input layer, convolutional layers, pooling layers, fully connected layers, and an output layer.

[0158] The convolutional layers are used to extract spatial and temporal related features of the data.

[0159] The pooling layers are used to reduce the dimensionality and computational amount of the data. Common pooling methods include max pooling and average pooling.

[0160] The fully connected layers map the extracted features to anomaly probabilities and prepare the output results.

[0161] The output layer outputs the anomaly detection results, usually using the softmax function for classification.

[0162] 3. Model training and optimization:

[0163] Use the labeled data to train the model.

[0164] In this embodiment, for the "stoma anomaly picture dataset", feature engineering is first performed to extract a large number of abnormal stoma feature images to form an abnormal stoma feature image set, and then feature learning is performed on the abnormal stoma feature images of different abnormal stoma types to enable the model to learn to recognize the abnormal stoma feature images of different abnormal stoma types.

[0165] Therefore, following the training steps of the convolutional neural network, after training and learning the abnormal stoma feature images of different abnormal stoma types, a stoma abnormality AI recognition model can be obtained, and the performance of the stoma abnormality AI recognition model can be verified.

[0166] During the training process, the backpropagation algorithm is used to optimize the model parameters so that the model can better fit the data.

[0167] To prevent overfitting, methods such as regularization and Dropout can be used for model regularization and optimization.

[0168] Monitor indicators such as the loss function and accuracy during the training process to adjust the training strategy in a timely manner.

[0169] 4. Model evaluation and tuning:

[0170] Use the validation set to evaluate the model, including indicators such as accuracy, recall rate, and F1 value.

[0171] Tune the model according to the evaluation results, such as adjusting the network structure, changing the learning rate, increasing the regularization strength, etc.

[0172] 5. Model deployment and application:

[0173] Deploy the trained model to the actual environment (the evaluation module in the background).

[0174] Perform anomaly detection on new stoma images and output the anomaly probability and possible anomaly types.

[0175] Take corresponding treatment measures according to the detection results, such as reminding the doctor to conduct further examinations or treatments.

[0176] 6. Continuous learning and improvement:

[0177] As new data accumulates, retrain and update the model regularly to improve its adaptability and accuracy.

[0178] Keep an eye on new technologies and methods and try to apply them to the stoma abnormality recognition model to further improve the performance of the model.

[0179] The above steps are for reference only, and specific implementation may need to be adjusted and optimized according to the actual situation.

[0180] As Figure 4 shown, it is a comparison picture of abnormal and normal stomas.

[0181] As an optional implementation of this application, optionally, after verifying whether the stoma abnormality AI recognition model meets the standard using the validation set, it further includes:

[0182] Prepare several normal stoma pictures;

[0183] Input the normal stoma pictures into the abnormal stoma AI recognition model to verify whether the abnormal stoma AI recognition model generates an image feature recognition reaction to the normal stoma pictures:

[0184] If it generates, re-collect the abnormal stoma pictures and re-train the model;

[0185] On the contrary, proceed to the next step.

[0186] In this verification, normal stoma pictures are also used to verify the model, to verify whether the model only recognizes abnormal stoma pictures and avoid interference of normal stoma pictures on the image recognition of the model.

[0187] As an optional implementation scheme of this application, optionally, the patient management system is further used for:

[0188] Retrieve the emergency education measures for different abnormal stomas from the MYSQL database according to a preset frequency, and send them to the patient-side APP to notify the patient to conduct stoma care education learning regularly.

[0189] The specific learning frequency can be set by the administrator.

[0190] The learning content can be the emergency education measures for different abnormal stomas. Professional learning videos can also be provided jointly by ostomy bag manufacturers and professional medical institutions:

[0191] The information security of patients needs to be guaranteed, and the qualification review of the medical institutions admitted needs to be strict;

[0192] For stoma recognition and abnormal condition recognition, biotechnology may be used to judge normal stomas, judge abnormal stoma data, and increase the credibility through a large number of recognition improvements;

[0193] The provision of relevant medical advice services for patients in need of diagnosis and treatment requires the support of medical institutions in various places;

[0194] The patient communication platform needs to be strictly reviewed to prevent behaviors that harm the interests of patients.

[0195] As an optional implementation scheme of this application, optionally, the patient-side APP is further used for:

[0196] Collect the learning logs of the patient's stoma care education learning and upload them to the background server, which is forwarded by the background server to the patient management system;

[0197] The patient management system is further configured to:

[0198] Bind the learning log of the patient under the corresponding patient's visit ID.

[0199] After each learning session of the patient, the mobile terminal can record the learning log of the patient and synchronize it to the background server, facilitating the background management of the patient's learning log.

[0200] As an optional implementation of the present application, optionally, the patient-side APP is further configured to:

[0201] The patient shares daily care information with the background server, and the background server forwards it to the patient management system;

[0202] The patient management system is further configured to:

[0203] Write the patient's daily care information into the patient management system and bind it under the corresponding patient's visit ID.

[0204] As an optional implementation of the present application, optionally, the doctor's PC terminal is further configured to:

[0205] Log in to the background server, enter the patient management system, view the stoma pictures and their abnormal assessment results of each patient, as well as the learning log and the daily care information; and,

[0206] Generate corresponding medical orders and write them into the patient management system, and the patient management system binds the medical orders under the corresponding patient's visit ID.

[0207] The doctor can view the various data / information of the patient after authorization, specifically referring to the doctor's information management and application system for patients in the existing hospital management system (such as the HIS system). The doctor can generate medical orders through his / her household management system and write them into the patient management system. After receiving the medical orders, the patient management system will forward them to the corresponding patient APP to remind the patient to view and execute the medical orders.

[0208] Therefore, through the patient uploading stoma pictures, the present application enables the background to perform online abnormal monitoring of the stoma and give corresponding emergency education measures, which can save time and medical staff resources, allowing the patient to perform stoma abnormal care at home and facilitating the patient's subsequent stoma care.

[0209] Obviously, those skilled in the art should understand that to implement all or part of the processes in the above embodiments, it can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above control embodiments. Those skilled in the art can understand that to implement all or part of the processes in the above embodiments, it can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above control embodiments. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above types of memories.

[0210] Embodiment 2

[0211] Based on the implementation principle of Embodiment 1, on the other hand, this application proposes a stoma abnormality evaluation method, which is implemented based on the stoma abnormality state online recognition and diagnosis evaluation system, and includes the following steps:

[0212] The patient takes a photo of the stoma site through the patient-side APP, collects the stoma picture and uploads it to the background server;

[0213] The background server performs intelligent evaluation of stoma abnormalities on the stoma picture to determine whether the stoma is abnormal:

[0214] If it is abnormal, an early warning notice of stoma abnormality is sent to the patient-side APP, and at the same time, emergency education measures for the corresponding abnormal stoma are issued;

[0215] If it is not abnormal, a notice of normal stoma is sent to the patient-side APP;

[0216] The doctor logs in to the background server through the doctor's PC terminal, follows up the stoma pictures and their abnormal evaluation results of each patient, determines whether to issue a doctor's order, and if a doctor's order is issued, forwards the doctor's order to the corresponding patient-side APP through the background server.

[0217] The above steps should be understood and implemented in combination with Embodiment 1.

[0218] Except for the above steps, other steps please refer to the interaction process in Embodiment 1.

[0219] Each module or step of the present invention described above can be implemented by a general computing system. They can be concentrated on a single computing system or distributed on a network composed of multiple computing systems. Optionally, they can be implemented by program code executable by the computing system. Thus, they can be stored in the storage system and executed by the computing system, or they can be separately fabricated into individual integrated circuit modules, or multiple modules or steps among them can be fabricated into a single integrated circuit module for implementation. In this way, the present invention is not limited to any specific combination of hardware and software.

[0220] Embodiment 3

[0221] As Figure 5 shown, further, on the other hand, the present application also proposes an electronic device, including:

[0222] A processor;

[0223] A memory for storing instructions executable by the processor;

[0224] Wherein, when the processor is configured to execute the executable instructions, it implements the stoma abnormality evaluation method described in Embodiment 2.

[0225] The electronic device according to an embodiment of the present disclosure includes a processor and a memory for storing instructions executable by the processor. Wherein, when the processor is configured to execute the executable instructions, it implements the stoma abnormality evaluation method described in Embodiment 2 above.

[0226] Here, it should be noted that the number of processors can be one or more. At the same time, in the electronic device according to an embodiment of the present disclosure, an input system and an output system can also be included. Wherein, the processor, the memory, the input system and the output system can be connected through a bus or in other ways, which is not specifically limited herein.

[0227] As a computer-readable storage medium, the memory can be used to store software programs, computer-executable programs and various modules, such as: the programs or modules corresponding to the stoma abnormality evaluation method according to an embodiment of the present disclosure. The processor executes various functional applications and data processing of the electronic device by running the software programs or modules stored in the memory.

[0228] The input system can be used to receive input numbers or signals. Wherein, the signal can be a key signal related to the user settings and function control of the device / terminal / server. The output system can include a display device such as a display screen.

[0229] The embodiments of the present disclosure have been described above. The above description is exemplary and not exhaustive, and is also not limited to the disclosed embodiments. Many modifications and variations are obvious to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The choice of terms used herein is intended to best explain the principles of the embodiments, the practical application, or the improvement of technologies in the market, or to enable other ordinary technical personnel in the art to understand the embodiments disclosed herein.

Claims

1. An online recognition and diagnosis and evaluation system for stoma abnormality, characterized in that: It includes the patient-side APP, the doctor's PC and the backend server, including: The patient-side APP is used for patients to log in to the backend server and upload their own stoma pictures; The backend server is used to perform an intelligent evaluation of stoma abnormality on the stoma image to determine whether the stoma is abnormal: If it is abnormal, an abnormal stoma warning notification will be sent to the patient APP, and emergency education measures corresponding to the abnormal stoma will be issued; If there is no abnormality, a normal stoma notification is sent to the patient APP; The doctor's PC is used for the doctor to log in to the backend server, follow up on the stoma pictures and abnormality assessment results of each patient, and determine whether to issue a medical order. If a medical order is issued, the medical order is forwarded to the corresponding patient-side APP through the backend server; The patient-side APP and the doctor's PC are respectively connected to the backend server for communication.

2. The system for online identification and diagnosis and evaluation of stoma abnormality according to claim 1, characterized in that: The backend server comprises: API interface, used for APP / PC access and data interaction; A medical care management system, used to manage the doctor's PC terminal; A patient management system, used to manage the patient-side APP and bind each patient's stoma picture and its abnormality assessment results to the corresponding patient's medical ID; MYSQL database, used to store stoma images of each patient and their abnormality assessment results, as well as to store the emergency education measures for different abnormal stomas; A stoma education module, used to store the mapping relationship between different abnormal stomas and the corresponding emergency education measures; The stoma abnormality intelligent assessment module is used to use the pre-deployed stoma abnormality AI recognition model to perform stoma abnormality recognition on the stoma image of the patient, and evaluate whether a preset abnormal stoma feature image appears in the stoma image of the patient: If so, the stoma abnormality identification result is output and the patient management system binds the stoma abnormality identification result to the corresponding patient's visit ID; Otherwise, the normal stoma recognition result is output and the patient management system binds the normal stoma recognition result to the corresponding patient's consultation ID.

3. The system for online identification and diagnosis and evaluation of stoma abnormality according to claim 2, characterized in that: The method for generating the stoma abnormality AI recognition model comprises: Collect several pictures of stoma abnormalities; Performing feature engineering on each of the abnormal stoma pictures, extracting abnormal stoma feature images on each of the abnormal stoma pictures, and forming an abnormal stoma feature image set; Dividing the abnormal stoma feature image set into a training set and a validation set according to a preset ratio; Importing the training set into a preset convolutional neural network model to perform deep learning training to generate the stoma abnormality AI recognition model; Using the verification set, verify whether the stoma abnormality AI recognition model meets the standards; If the verification is successful, the stoma abnormality AI recognition model is deployed on the background server; If the verification fails to meet the standards, feature engineering will be repeated and the model will be retrained.

4. The system for online identification and diagnosis and evaluation of stoma abnormality according to claim 3, characterized in that: After using the verification set to verify whether the stoma abnormality AI recognition model meets the standard, the method further includes: Prepare several normal stoma pictures; Input the normal stoma picture into the stoma abnormality AI recognition model to verify whether the stoma abnormality AI recognition model generates an image feature recognition response to the normal stoma picture: If so, recollect the stoma abnormality picture and retrain the model; Otherwise, proceed to the next step.

5. The system for online identification and diagnosis and evaluation of stoma abnormality according to claim 2, characterized in that: The patient management system is also used for: According to the preset frequency, the emergency education measures for different abnormal stomas are retrieved from the MYSQL database and sent to the patient-side APP to notify the patient to conduct stoma care education and learning regularly.

6. The system for online identification and diagnosis and evaluation of stoma abnormality according to claim 5, characterized in that: The patient-side APP is also used for: Collecting the patient's learning log of stoma care education and learning, and uploading it to the backend server, which is then forwarded to the patient management system; The patient management system is also used for: Bind the patient's learning log to the corresponding patient's visit ID.

7. The system for online identification and diagnosis and evaluation of stoma abnormality according to claim 2, characterized in that: The patient-side APP is also used for: The patient shares daily care information with the backend server, which is forwarded to the patient management system by the backend server; The patient management system is also used for: The patient's daily care information is written into the patient management system and bound to the corresponding patient's visit ID.

8. The system for online identification and diagnosis and evaluation of stoma abnormality according to claim 7, characterized in that: The doctor's PC terminal is also used for: Log in to the backend server, enter the patient management system, view each patient's stoma picture and abnormality assessment results, as well as the learning log and the daily care information; as well as, Generate a corresponding medical order and write it into the patient management system, which binds the medical order to the corresponding patient's consultation ID.

9. A stoma abnormality assessment method, implemented based on the stoma abnormality state online identification and diagnosis assessment system according to any one of claims 1 to 8, characterized in that: The steps include: The patient takes a photo of the stoma through the patient-side APP, collects the stoma image and uploads it to the backend server; The backend server performs an intelligent stoma abnormality assessment on the stoma image to determine whether the stoma is abnormal: If it is abnormal, an abnormal stoma warning notification will be sent to the patient APP, and emergency education measures corresponding to the abnormal stoma will be issued; If there is no abnormality, a normal stoma notification is sent to the patient APP; The doctor logs in to the backend server through the doctor's PC, follows up on the stoma pictures and abnormality assessment results of each patient, and determines whether to issue medical advice. If medical advice is issued, the medical advice is forwarded to the corresponding patient-side APP through the backend server.

10. An electronic device, characterized in that include: processor; a memory for storing processor-executable instructions; Wherein, the processor is configured to implement the stoma abnormality assessment method according to any one of claims 1 to 8 when executing the executable instructions.

Images