Radiotherapy patient rehabilitation monitoring system based on deep learning
By designing a radiotherapy patient rehabilitation monitoring system based on deep learning, the problems of limitations of data collection and low evaluation accuracy in the existing system are solved, and multi-dimensional data collection and dynamic monitoring are realized, which improves patient participation and rehabilitation effect.
Patent Information
- Application Number
- CN202510462616.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-14
- Publication Date
- 2025-06-13
AI Technical Summary
The existing radiotherapy patient rehabilitation monitoring system has great limitations in data collection, lack of real-time performance, and insufficient data preprocessing, resulting in low data evaluation and analysis accuracy, low patient participation, and the inability to dynamically monitor the rehabilitation process.
A radiotherapy patient rehabilitation monitoring system based on deep learning is designed, including a patient data acquisition terminal, a remote data transmission terminal, a data analysis processing terminal and a rehabilitation monitoring feedback terminal. The system collects vital sign information in real time through intelligent wearable devices, combines the radiotherapy response information and psychological status information uploaded by the patient independently, performs data preprocessing and deep learning analysis, dynamically evaluates the patient's recovery status, and provides personalized rehabilitation plans and remote rehabilitation guidance.
Multi-directional collection and dynamic monitoring of data are realized, the accuracy of data evaluation and patient participation are improved, the response time of postoperative rehabilitation intervention is shortened, and the postoperative rehabilitation effect of patients is significantly improved.
Smart Images
Figure CN120130971A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of rehabilitation monitoring, and particularly to a radiotherapy patient rehabilitation monitoring system based on deep learning. Background Art
[0002] With the continuous development of modern medical technology, the medical practice in the field of radiotherapy has achieved rapid development and progress. As an important treatment technology, radiotherapy has been widely applied to all stages of cancer treatment. Radiotherapy is one of the three major treatment methods for malignant tumors, and about 70% of cancer patients need radiotherapy. Moreover, due to the long treatment course and large side effects of radiotherapy, even if the patient's tumor has been cured after surgery, continuous rehabilitation treatment is still required, which is called the tumor rehabilitation period. Specifically, it refers to the period when the cancer has been clinically cured after treatment by means such as surgery, chemotherapy, and radiotherapy, and the patient no longer undergoes treatment but is recuperating, that is, physical rehabilitation in the biological sense. Therefore, in order to improve the rehabilitation quality of patients, it is usually necessary to effectively monitor the postoperative rehabilitation situation of patients.
[0003] Deep learning models are a very powerful class of machine learning models. In the past few years, they have achieved remarkable success in many fields, such as computer vision, natural language processing, and speech recognition. The core of deep learning models is the neural network, and the number of its layers is determined by the complexity of the training data. Through a large amount of data input and preprocessing, deep learning models can automatically learn features and rules and obtain very good prediction results.
[0004] At present, the research on applying deep learning models to the rehabilitation monitoring of radiotherapy patients is not deep enough. Most of the existing radiotherapy patient rehabilitation monitoring systems have single functions, cannot comprehensively and effectively collect the relevant postoperative data of patients, lack real-time data collection, resulting in large limitations in data collection, and there is no effective preprocessing after data collection, resulting in incomplete and inaccurate problems, thus having an adverse impact on the accuracy of subsequent data evaluation and analysis. Moreover, the participation of patients in the entire monitoring process is relatively low, which to a certain extent affects the enthusiasm and rehabilitation effect of patients. At the same time, it is also unable to dynamically monitor and evaluate the rehabilitation process of patients, and the rehabilitation evaluation accuracy is relatively low. Therefore, the present invention proposes a radiotherapy patient rehabilitation monitoring system based on deep learning to solve the problems existing in the prior art. Summary of the Invention
[0005] Aiming at the above problems, the purpose of the present invention is to propose a radiotherapy patient rehabilitation monitoring system based on deep learning to solve the problems of large limitations in data collection of the existing radiotherapy patient rehabilitation monitoring system, no effective preprocessing after data collection, and inability to dynamically monitor and evaluate the rehabilitation process of patients.
[0006] To achieve the object of the present invention, the present invention is implemented through the following technical solutions: A radiotherapy patient rehabilitation monitoring system based on deep learning, comprising a patient data acquisition terminal, a remote data transmission terminal, a data analysis and processing terminal, and a rehabilitation monitoring feedback terminal. The patient data acquisition terminal includes a vital sign monitoring module for collecting the vital sign information of the patient, a radiotherapy reaction detection module for collecting the radiotherapy reaction information of the patient, and a psychological state collection module for collecting the psychological state information of the patient;
[0007] The remote data transmission terminal includes a data sending module carried on the patient data acquisition terminal, a data cloud backup module accessing the Internet, and a data receiving module carried on the data analysis and processing terminal. Wireless data transmission is carried out between the patient data acquisition terminal and the data analysis and processing terminal through the remote data transmission terminal;
[0008] The data analysis and processing terminal includes a data preprocessing module for preprocessing the received patient data, a data analysis module for performing deep learning analysis on the preprocessed patient data, and a data evaluation module for evaluating the rehabilitation status of the patient according to the data analysis result. The data analysis and processing terminal has a growing database. The data preprocessing module includes a data cleaning unit, a missing data processing unit, and a data standardization unit;
[0009] The rehabilitation monitoring feedback terminal includes a hierarchical warning module wirelessly connected to the data analysis and processing terminal and a rehabilitation guidance module wirelessly connected to the patient data acquisition terminal. The hierarchical warning module grades the patient's rehabilitation situation and formulates a personalized rehabilitation plan. The rehabilitation guidance module provides remote rehabilitation guidance for the patient according to the personalized rehabilitation plan.
[0010] A further improvement lies in that: The vital sign monitoring module uses intelligent wearable devices to monitor and collect the vital sign information of the patient in real time. The radiotherapy reaction detection module receives the skin reaction pictures independently uploaded by the patient and the sensory information reported orally by the patient. The psychological state collection module asks the patient psychological state evaluation questions in the form of a questionnaire and evaluates the psychological state information of the patient according to the patient's answers.
[0011] A further improvement lies in that: The data sending module classifies and compresses the information data collected by the patient data acquisition terminal and sends it to the data receiving module. The data receiving module decompresses the received information data packet and transmits it to the data preprocessing module. The data cloud backup module stores the collected information data in the cloud server in real time for backup.
[0012] A further improvement lies in that: the data cleaning unit uses wavelet transform technology to remove high-frequency noise in the information data and retain the effective signals. At the same time, it detects abnormal patterns in the information data through an autoencoder and corrects the outliers according to the clinical background.
[0013] A further improvement lies in that: the missing value handling unit fills the data holes by selecting corresponding data filling methods based on the data characteristics, and the data standardization unit uses the Min-Max normalization technology to compress the data into the interval [0, 1] to unify the data scale.
[0014] A further improvement lies in that: the data analysis module includes a deep learning module for extracting key features in the information data through a deep learning model and a comparison and recognition module for identifying the patient's rehabilitation stage. The deep learning module constructs a deep learning model based on a convolutional neural network, takes the preprocessed data as the model input, automatically analyzes the data using the deep learning model and extracts the key features therein. The comparison and recognition module compares the extracted key features with the standard features and historical features stored in the growth database and identifies the patient's rehabilitation stage according to the comparison results.
[0015] A further improvement lies in that: the standard features stored in the growth database are the ideal rehabilitation states of patients after radiotherapy. The data evaluation module evaluates the patient's rehabilitation progress according to the recognition result of the patient's rehabilitation stage and generates a rehabilitation monitoring report for the patient.
[0016] A further improvement lies in that: the grading and early warning module grades the patient's rehabilitation condition according to the patient's rehabilitation progress and formulates a personalized rehabilitation plan for the patient according to the grading of the rehabilitation condition. The rehabilitation guidance module uses the telemedicine mode to provide professional rehabilitation guidance for the patient.
[0017] The beneficial effects of the present invention are as follows: by integrating objective sensor detection data and subjective patient-collected data, the present invention dynamically associates the physical parameters of postoperative rehabilitation with the patient's physiological responses, realizes multi-faceted collection of data, improves the comprehensiveness of subsequent data evaluation, and also improves the patient's participation, ensures the patient's enthusiasm, and adopts a hybrid data cleaning scheme of wavelet transform plus autoencoder, which can effectively preprocess the data, further improve the efficiency of subsequent data processing and analysis, and realizes accurate dynamic judgment of the rehabilitation process through deep learning, significantly improving the rehabilitation evaluation accuracy. The rehabilitation monitoring feedback terminal provides remote rehabilitation guidance for the patient, realizes real-time intervention through telemedicine, forms a multi-terminal collaborative architecture and a closed-loop structure of monitoring, analysis and feedback, shortens the response time of postoperative rehabilitation intervention, and has a positive impact on the postoperative rehabilitation effect of the patient. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings.
[0019] Figure 1 It is a schematic structural diagram of the radiotherapy patient rehabilitation monitoring system based on deep learning of the present invention. Detailed implementation manners
[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.
[0021] Deep Learning (DL for short) is a branch of machine learning. It simulates the neural network structure of the human brain and uses a multi-layer neural network model to learn high-level features of data. Deep learning models usually contain multiple hidden layers, and each layer transforms the input data to extract higher-level abstract features. As a powerful machine learning method, deep learning has achieved remarkable achievements in many fields such as image recognition, speech recognition, and natural language processing.
[0022] The rehabilitation monitoring of radiotherapy patients is a comprehensive process, which refers to the monitoring and evaluation of the patient's physical condition, treatment response, quality of life, etc. during and after radiotherapy. This kind of monitoring helps to timely detect and handle the adverse reactions caused by radiotherapy, optimize the treatment plan, and improve the patient's quality of life.
[0023] See Figure 1 , this embodiment provides a radiotherapy patient rehabilitation monitoring system based on deep learning. The system consists of a patient data acquisition terminal for collecting postoperative information of radiotherapy patients, a remote data transmission terminal for remotely wirelessly transmitting the collected data, a data analysis and processing terminal for processing and analyzing the remotely transmitted data, and a rehabilitation monitoring feedback terminal for making corresponding feedback work according to the data processing and analysis results. Among them, the patient data acquisition terminal is wirelessly connected to the data analysis and processing terminal through the remote data transmission terminal for data transmission, and the rehabilitation monitoring feedback terminal is wirelessly connected to the patient data acquisition terminal and the data analysis and processing terminal respectively.
[0024] The patient data acquisition terminal in this embodiment consists of a vital sign monitoring module, a radiotherapy reaction detection module, and a psychological state collection module. The vital sign monitoring module is used to collect the patient's vital sign information. The radiotherapy reaction detection module is used to collect the patient's radiotherapy reaction information. The psychological state collection module is used to collect the patient's psychological state information;
[0025] The remote data transmission terminal consists of a data sending module, a data cloud backup module, and a data receiving module. The data sending module is carried on the patient data acquisition terminal and compresses and packages the collected data. The data cloud backup module accesses the Internet and performs cloud backup storage on the collected data. The data receiving module is carried on the data analysis and processing terminal and receives the data remotely sent from the data sending module;
[0026] The data analysis and processing terminal consists of a data preprocessing module, a data analysis module, and a data evaluation module. The data preprocessing module is used to preprocess the received patient data. The data analysis module is used to perform in-depth learning analysis on the preprocessed patient data. The data evaluation module is used to evaluate the patient's rehabilitation status based on the data analysis results. The data preprocessing module includes a data cleaning unit for eliminating noise and outliers, a missing value processing unit for filling in missing values, and a data standardization unit for standardizing the data;
[0027] The rehabilitation monitoring and feedback terminal consists of a hierarchical warning module and a rehabilitation guidance module. The hierarchical warning module is wirelessly connected to the data analysis and processing terminal and is used to classify the patient's rehabilitation situation and formulate a personalized rehabilitation plan. The rehabilitation guidance module is wirelessly connected to the patient data acquisition terminal and provides remote rehabilitation guidance for the patient according to the personalized rehabilitation plan formulated by the hierarchical warning module.
[0028] The vital sign monitoring module uses intelligent wearable devices to monitor and collect the patient's vital sign information in real time. The patient wears intelligent wearable devices including smart bracelets. The intelligent wearable device also has voice recording and broadcasting functions. The collected information includes heart rate, blood pressure, respiratory rate, and body temperature. Wearable data monitoring can improve the real-time and convenience of data collection and, to a certain extent, improve the accuracy of subsequent data processing and analysis;
[0029] The radiotherapy reaction detection module receives the skin reaction pictures independently uploaded by the patient and the sensory information including fatigue, nausea, and pain reported orally by the patient, realizes the indirect detection of the patient's physical reaction after radiotherapy, and at the same time improves the patient's participation in postoperative rehabilitation monitoring work;
[0030] The mental state collection module asks the patient mental state assessment questions in the form of a Q&A, and evaluates the patient's mental state information including anxiety and depression based on the patient's answers. It asks the patient questions by voice according to the verified mental assessment scale, quickly screens the patient's states such as anxiety and depression, and realizes the collection of the patient's mental health state data.
[0031] The data sending module classifies and compresses the information data collected by the patient data collection terminal and sends it to the data receiving module. Data classification facilitates subsequent data analysis, and compression improves data transmission speed. The data receiving module decompresses the received information data packet and transmits it to the data preprocessing module. The data cloud backup module stores the collected information data in the cloud server in real time for subsequent traceability.
[0032] The data cleaning unit uses wavelet transform technology to remove the high-frequency noise in the information data and retain the effective signal. At the same time, it detects the abnormal patterns in the information data through an autoencoder, and judges the rationality of the outliers according to the clinical background, and corrects the outliers according to the rationality.
[0033] The missing data processing unit selects the corresponding data filling method based on the data characteristics to fill the data loopholes. For physiological time series data, time series prediction technology is used to fill the missing data. For scale data, multiple imputation is used to fill the missing data.
[0034] The data standardization unit uses the Min-Max normalization technology to compress the data into the interval [0,1] to unify the data scale and use it as the input of the subsequent deep learning model.
[0035] The data analysis module consists of a deep learning module and a comparison and recognition module and comes with a growing database. The growing database combines incremental learning, federated learning and knowledge graph verification technology to dynamically update the standard features, and realizes the continuous optimization of the standard features under the premise of protecting privacy. The deep learning module constructs a deep learning model based on a convolutional neural network, inputs the preprocessed data into the constructed deep learning model, and automatically analyzes the patient data using the deep learning model to extract the key features. The comparison and recognition module compares the extracted key features with the standard features and historical features stored in the growing database, and identifies the patient's rehabilitation stage according to the comparison results. The standard features stored in the growing database are the ideal rehabilitation state after the patient's radiotherapy, which are pre-entered by the medical staff according to the patient's radiotherapy situation. The data evaluation module evaluates the patient's rehabilitation progress according to the recognition result of the patient's rehabilitation stage and generates a rehabilitation monitoring report for the patient.
[0036] The grading early warning module grades the patient's rehabilitation status according to the patient's rehabilitation progress and in combination with clinical guidelines (CTCAE adverse reaction grading), and formulates a personalized rehabilitation plan for the patient based on the graded rehabilitation status. The rehabilitation guidance module uses the telemedicine mode to provide professional rehabilitation guidance to the patient, guiding the patient in exercise training, nutritional supplement adjustment, and psychological intervention and counseling.
[0037] When the radiotherapy patient rehabilitation monitoring system based on deep learning is actually used, first, the vital sign monitoring module, radiotherapy reaction detection module, and psychological state collection module in the patient data acquisition terminal are used to collect the patient's vital sign information, radiotherapy reaction information, and psychological state information respectively. Then, the collected data are sent to the data analysis and processing terminal through the remote data transmission terminal. Next, the data preprocessing module, data analysis module, and data evaluation module in the data analysis and processing terminal are used to preprocess and analyze the received data successively, and evaluate the patient's rehabilitation status. Finally, a personalized rehabilitation plan is formulated through the rehabilitation monitoring feedback terminal, and real-time remote rehabilitation guidance is provided to the patient according to the plan.
[0038] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A radiotherapy patient rehabilitation monitoring system based on deep learning, comprising a patient data acquisition terminal, a remote data transmission terminal, a data analysis and processing terminal, and a rehabilitation monitoring feedback terminal, characterized in that: The patient data collection terminal includes a vital sign monitoring module for collecting vital sign information of the patient, a radiotherapy response detection module for collecting radiotherapy response information of the patient, and a psychological state collection module for collecting psychological state information of the patient; The remote data transmission terminal includes a data sending module mounted on the patient data collection terminal, a data cloud backup module connected to the Internet, and a data receiving module mounted on the data analysis and processing terminal. Wireless data transmission is performed between the patient data collection terminal and the data analysis and processing terminal via the remote data transmission terminal; The data analysis and processing terminal includes a data preprocessing module for preprocessing received patient data, a data analysis module for performing deep learning analysis on the preprocessed patient data, and a data evaluation module for evaluating the patient's rehabilitation status according to the data analysis results. The data analysis and processing terminal has a growing database, and the data preprocessing module includes a data cleaning unit, a missing processing unit, and a data standardization unit. The rehabilitation monitoring feedback terminal includes a graded warning module wirelessly connected to the data analysis and processing terminal and a rehabilitation guidance module wirelessly connected to the patient data acquisition terminal. The graded warning module grades the patient's rehabilitation status and formulates a personalized rehabilitation plan, and the rehabilitation guidance module provides remote rehabilitation guidance to the patient according to the personalized rehabilitation plan.
2. According to claim 1, a radiotherapy patient rehabilitation monitoring system based on deep learning is characterized in that: The vital signs monitoring module uses smart wearable devices to monitor and collect the patient's vital signs information in real time. The radiotherapy reaction detection module receives skin reaction pictures uploaded by the patient and sensory information self-reported by the patient. The psychological state collection module uses a question-and-answer method to ask the patient psychological state assessment questions and assesses the patient's psychological state information based on the patient's answers.
3. The radiotherapy patient rehabilitation monitoring system based on deep learning according to claim 1, characterized in that: The data sending module classifies the information data collected by the patient data acquisition terminal, compresses and packages them and sends them to the data receiving module. The data receiving module decompresses the received information data packets and transmits them to the data preprocessing module. The data cloud backup module backs up the collected information data in real time and stores them in the cloud server.
4. The radiotherapy patient rehabilitation monitoring system based on deep learning according to claim 1, characterized in that: The data cleaning unit uses wavelet transform technology to remove high-frequency noise in the information data and retain valid signals. At the same time, it detects abnormal patterns in the information data through an automatic encoder and corrects abnormal values according to clinical background.
5. The radiotherapy patient rehabilitation monitoring system based on deep learning according to claim 1, characterized in that: The missing processing unit selects a corresponding data filling method based on data characteristics to fill data gaps, and the data standardization unit uses Min-Max normalization technology to compress the data to the [0,1] interval to unify the data scale.
6. The radiotherapy patient rehabilitation monitoring system based on deep learning according to claim 1, characterized in that: The data analysis module includes a deep learning module that extracts key features from information data through a deep learning model and a comparative identification module for identifying the patient's rehabilitation stage. The deep learning module constructs a deep learning model based on a convolutional neural network, and uses the preprocessed data as a model input. The deep learning model is used to automatically analyze the data and extract key features therein. The comparative identification module compares the extracted key features with the standard features and historical features stored in the growing database, and identifies the patient's rehabilitation stage based on the comparison results.
7. A radiotherapy patient rehabilitation monitoring system based on deep learning according to claim 6, characterized in that: The standard features stored in the growing database are the idealized rehabilitation status of the patient after radiotherapy. The data evaluation module evaluates the patient's rehabilitation progress based on the identification result of the patient's rehabilitation stage and generates a rehabilitation monitoring report for the patient.
8. The radiotherapy patient rehabilitation monitoring system based on deep learning according to claim 1, characterized in that: The graded warning module grades the patient's rehabilitation status according to the patient's rehabilitation progress, and formulates a personalized rehabilitation plan for the patient according to the rehabilitation status classification. The rehabilitation guidance module adopts a telemedicine mode to provide professional rehabilitation guidance to the patient.