Old people health status assessment data processing system based on multi-modal data

Abstract

The invention relates to the technical field of data processing, and discloses an old people health state assessment data processing system based on multi-modal data, and the system comprises a medical data integration module which obtains electronic medical record data and a medical examination report through an FHIR interface, and extracts a structured health index; the cross-modal causal fusion processing module is used for fusing the monitoring data and the medical text through an image, text and image interlayer architecture; the health state evolution modeling module is used for mapping the health feature vectors into physiological function, cognitive level and athletic ability three-dimensional state indexes; an evaluation report backtracking module; and a decision output module. Through an image, text and image interlayer architecture, deep semantic fusion of multi-modal features is realized under the constraint of medical pathology rules, feature weight adaptive distribution is dynamically guided based on a medical causal atlas, a high-dimensional fusion vector retaining key pathology information is generated, the semantic integration ability of health data is remarkably improved, and the health data fusion efficiency is improved. And the reliability of discrimination and decision making is improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a data processing system for evaluating the health status of elderly people based on multimodal data. Background Art

[0002] Faced with the accelerating trend of global population aging, complex health problems such as multiple diseases and dynamic evolution of health status among the elderly are becoming increasingly prominent. Real-time and accurate health status assessment has become a key link in achieving proactive intervention and delaying functional decline. In response to the health management needs of the elderly, it is necessary to build a scientific and quantitative assessment system to achieve early warning and intervention of health risks. In this context, by integrating multi-dimensional health information flows, the scientificity, timeliness and effectiveness of elderly health management can be improved. Traditional multimodal health assessment solutions mostly use feature splicing or attention-weighted fusion to process image and text data.

[0003] However, in current technologies, the lack of pathological rule-guided feature allocation between image features and text semantics reduces clinical interpretability, resulting in reduced accuracy of health status assessment and reduced decision reliability. Summary of the Invention

[0004] In response to the shortcomings of the existing technology, the present invention provides a data processing system for the health status assessment of the elderly based on multimodal data, which solves the problem that the feature allocation guided by pathological rules in the fusion of image features and text semantics reduces clinical interpretability, resulting in reduced accuracy of health status assessment and decision reliability.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a data processing system for evaluating the health status of the elderly based on multimodal data, comprising: Medical data integration module, which obtains electronic medical record data and medical examination reports through the FHIR interface and extracts structured health indicators; A cross-modal causal fusion processing module, which fuses monitoring data and medical text through image, text, and image sandwich architecture to generate multi-dimensional health feature vectors under the constraints of medical causal graphs; The health status evolution modeling module maps health feature vectors into three-dimensional status indicators of physiological function, cognitive level, and motor ability, and calculates the risk time trajectory; Retrospective assessment report module, which analyzes the causal chain of historical health events and generates dynamic assessment reports based on the risk time trajectory; The decision output module is used to distribute the evaluation results to medical terminals and family devices.

[0006] Preferably, the cross-modal causal fusion processing module includes: an image feature extraction unit configured to process the medical image data through a three-dimensional convolutional network and output an anatomical structure feature matrix; A text encoding unit, configured to encode medical record text into semantic vectors using the BioBERT model; The causal constraint fusion unit divides the image features into front and back sub-matrices and embeds them into text vectors to form a sandwich structure. It applies the disease development path constraint feature weight distribution in the medical causal graph and outputs a multi-dimensional health feature vector.

[0007] Preferably, the disease development path constraints in the causal constraint fusion unit include: Path feature mapping, extracting the disease development path corresponding to the current diagnosis from the medical causal graph , the path node Encoded as a path feature vector ,in, It is a 300-dimensional embedding vector based on node types, including symptoms, diseases, and complications. The data source is the medical knowledge graph mapped by the ICD-11 code library; Attention constraint mechanism, calculating text feature vector and The cosine similarity is used as the weight constraint factor: ; in, The 768-dimensional medical record text vector output by BioBERT, is the dynamic weight factor; Weighted fusion, reorganized feature matrix:

[0008] in, is the first 1 / 3 slice of the image feature, The last 2 / 3 slice of the image feature. is the Hadamard product, and the constraint is enforced when End-stage heart failure .

[0009] Preferably, the health status evolution modeling module includes: A state mapping unit converts the multidimensional health feature vector into a physiological state index, a cognitive state index, and a motor ability index through linear transformation; A degradation analysis unit analyzes the changing trends of various state indices using an aging trajectory model that includes degradation patterns in three dimensions: physiological, cognitive, and motor. The risk calculation unit monitors the rate of decline of the exercise capacity index. When the continuous monitoring shows that the exercise capacity index accelerates and drops beyond the preset threshold, it generates a fall risk time window warning and establishes an association with the extracted recent fluctuating body movement data.

[0010] Preferably, the degradation analysis unit includes: Trend extraction: using convolutional neural networks to analyze the temporal variation patterns of various health indices and extract characteristic fragments of degradation in physiological, cognitive, and motor dimensions; Aging trajectory prediction, based on the extracted degradation features, predicts the health status evolution curve in the next 90 days through a linear mixed effects model.

[0011] Preferably, the retrospective evaluation report module includes: The timeline construction unit uses the risk time trajectory and integrates health indicators to establish a unified time base for the evolution of health status. The causal chain analysis unit analyzes the causal relationship between health events based on the medical causal graph and calculates the contribution of each event to the current health risk; The visual reporting engine integrates the risk time trajectory with the causal analysis results to generate a dynamic assessment report that includes a timeline heat map, event correlation network diagram, and risk prediction curve.

[0012] Preferably, the causal chain analysis unit includes: Causal graph construction is based on the medical causal graph, which generates a causal impact network diagram with the current health event as the root node; Contribution quantification: The contribution of each health event to the current risk is calculated using the following formula:

[0013] in, is the preset weight of the disease path in the causal graph, The risk change rate provided to the health state evolution modeling module, is the time interval from the event to the present, is a critical event indicator, i.e., 1 if it is a critical pathological stage, and 0 otherwise; Causal chain generation is to screen event nodes with a contribution greater than 0.15 and construct a multi-level causal chain with weight identification.

[0014] Preferably, the decision output module includes: A medical report distribution unit sends an assessment report to a medical terminal, wherein the report includes a causal chain of health events and a risk time trajectory; The family warning distribution unit extracts key risk information from the assessment report, generates simplified warning notifications and pushes them to family devices.

[0015] Preferably, the family warning distribution unit includes: When the risk level exceeds the threshold of health indicators, a voice announcement will be made; Adjust the voice broadcast volume according to the user's age; Display the risk time trajectory graph on the notification interface.

[0016] A data processing method for evaluating the health status of the elderly based on multimodal data, the method comprising the following steps: S1. Obtain electronic medical record data and medical examination reports through the FHIR interface and extract structured health indicators; S2, using image, text and image sandwich architecture to fuse monitoring data and medical text, generating multi-dimensional health feature vectors under the constraints of medical causal graph; S3, mapping the health feature vector into three-dimensional state indicators of physiological function, cognitive level and motor ability, and calculating the risk time trajectory; S4. Analyze the causal chain of historical health events and generate a dynamic assessment report based on the risk time trajectory; S5. Distribute the evaluation results to medical terminals and family members’ devices.

[0017] The present invention provides a data processing system for assessing the health status of the elderly based on multimodal data. It has the following beneficial effects: 1. This invention uses an image, text, and image sandwich architecture to achieve deep semantic fusion of multimodal features under the constraints of medical pathology rules. Based on the medical causal graph, it dynamically guides the adaptive allocation of feature weights to generate a high-dimensional fusion vector that retains key pathology information, significantly improving the semantic integration capability of health data and the reliability of judgment and decision-making.

[0018] 2. This invention establishes a quantitative indicator system for physiological function, cognitive level and motor ability, and combines it with the law of temporal degradation to build a personalized aging trajectory prediction model. Through the linkage analysis of multi-dimensional status indicators, it accurately captures the critical points of health risks and realizes a paradigm shift from static assessment to dynamic prediction.

[0019] 3. The present invention constructs a quantifiable contribution analysis model through a medical causal graph, integrates risk time trajectories with historical health events, and generates an assessment report with causal traceability through dynamic calculation of event weights, time decay factors, and key pathological markers, thereby enhancing the clinical reliability of the assessment report. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is an architecture diagram of a data processing system for evaluating the health status of elderly people based on multimodal data according to the present invention; Figure 2 This is a flow chart of the data processing method for evaluating the health status of the elderly based on multimodal data of the present invention. DETAILED DESCRIPTION

[0021] The following will clearly and completely describe the technical solution of the present invention in conjunction with the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0022] Please see the attached Figure 1 The embodiment of the present invention provides a data processing system for evaluating the health status of the elderly based on multimodal data, comprising: Medical data integration module, which obtains electronic medical record data and medical examination reports through the FHIR interface and extracts structured health indicators; A cross-modal causal fusion processing module, which fuses monitoring data and medical text through image, text, and image sandwich architecture to generate multi-dimensional health feature vectors under the constraints of medical causal graphs; The health status evolution modeling module maps health feature vectors into three-dimensional status indicators of physiological function, cognitive level, and motor ability, and calculates the risk time trajectory; Retrospective assessment report module, which analyzes the causal chain of historical health events and generates dynamic assessment reports based on the risk time trajectory; The decision output module is used to distribute the evaluation results to medical terminals and family devices.

[0023] Specifically, the medical data integration module integrates multi-source heterogeneous medical information through the standardized FHIR interface, converts electronic medical records and medical examination reports into structured health indicators, and provides clinical data input support in a unified format for subsequent processing modules; The cross-modal causal fusion processing module uses a sandwich architecture of images, text, and images to achieve deep feature fusion of multimodal medical data under the guidance of medical causal graphs, generating a multi-dimensional comprehensive feature vector representing health status, providing a unified high-quality feature expression for subsequent health status evolution modeling; The health status evolution modeling module uses a dynamic quantitative model to map multi-dimensional health feature vectors to a three-dimensional indicator system of physiological function, cognitive level, and motor ability. It constructs a health status trajectory based on time series evolution, predicts individualized risk development paths, and generates a visual risk time evolution curve, thus achieving continuous dynamic monitoring of the health status of the elderly and risk early warning. The retrospective assessment report module quantitatively analyzes the causal association paths between historical health events, combines the risk time trajectory output by health status evolution modeling, and integrates time series characteristics with causal reasoning results to generate an interactive assessment report. This enables visual tracking of the health risk evolution process and provides a comprehensive assessment output with temporal continuity and causal interpretability for medical decision-making. The decision output module distributes dynamic assessment reports to medical terminals and family devices, enabling efficient delivery of assessment conclusions and providing different user groups with information output channels adapted to their roles.

[0024] The cross-modal causal fusion processing module includes: an image feature extraction unit configured to process the medical image data through a three-dimensional convolutional network and output an anatomical structure feature matrix; A text encoding unit, configured to encode medical record text into semantic vectors using the BioBERT model; The causal constraint fusion unit divides the image features into front and back sub-matrices and embeds them into text vectors to form a sandwich structure. It applies the disease development path constraint feature weight distribution in the medical causal graph and outputs a multi-dimensional health feature vector.

[0025] Specifically, the image feature extraction unit analyzes the spatial structural information of medical images through a three-dimensional convolutional neural network, extracts the morphological and texture features of anatomical tissues from multimodal medical images, and converts the original image data into a high-dimensional anatomical structure representation matrix; The text encoding unit uses the BioBERT pre-trained language model to parse the clinical semantics of medical record texts. Through medical entity recognition and relationship extraction technology, it encodes unstructured text into a vector that retains medical semantics. The causal constraint fusion unit constructs a sandwich fusion structure of image features and text semantics based on the disease development path defined by the medical causal graph: it decouples the image features into different subspaces, embeds the text semantic vector associated with the pathology, and outputs a multi-dimensional health feature vector through dynamic weight distribution.

[0026] The disease development path constraints in the causal constraint fusion unit include: Path feature mapping, extracting the disease development path corresponding to the current diagnosis from the medical causal graph , the path node Encoded as a path feature vector ,in, It is a 300-dimensional embedding vector based on node types, including symptoms, diseases, and complications. The data source is the medical knowledge graph mapped by the ICD-11 code library; Attention constraint mechanism, calculating text feature vector and The cosine similarity is used as the weight constraint factor: ; in, The 768-dimensional medical record text vector output by BioBERT, is the dynamic weight factor; Weighted fusion, reorganized feature matrix:

[0027] in, is the first 1 / 3 slice of the image feature, The last 2 / 3 slice of the image feature. is the Hadamard product, and the constraint is enforced when End-stage heart failure .

[0028] Specifically, path feature mapping extracts the currently diagnosed disease development path sequence from the medical causal graph, maps the symptom, disease, and complication nodes in the path into embedding vectors, and establishes a continuous feature representation of the disease evolution process, providing structured knowledge guidance for subsequent feature fusion. The attention constraint mechanism calculates the semantic similarity between text features and path features, generates dynamic weight factors that reflect clinical semantic relevance, and automatically aligns medical record text descriptions with disease development pathways, ensuring that key pathology features receive reasonable weight distribution during the fusion process. The weighted fusion unit, whose sandwich reconstruction strategy decouples image features into front and back feature subspaces, and embeds the constrained text semantic vector in between, realizes local weighting of the posterior segment features through the Hadamard product, enforces the key feature retention mechanism when diagnosing specific pathological stages, and finally outputs multidimensional fusion features with clinical interpretability.

[0029] The health status evolution modeling module includes: A state mapping unit converts the multidimensional health feature vector into a physiological state index, a cognitive state index, and a motor ability index through linear transformation; The degradation analysis unit analyzes the changing trends of various state indices through the aging trajectory model, which includes degradation patterns in three dimensions: physiological, cognitive, and motor. The risk calculation unit monitors the rate of decline of the exercise capacity index. When the continuous monitoring shows that the exercise capacity index accelerates and drops beyond the preset threshold, it generates a fall risk time window warning and establishes an association with the extracted recent fluctuating body movement data.

[0030] Specifically, the state mapping unit projects the multidimensional health feature vector into the three-dimensional index space of physiological function, cognitive function and motor ability through linear transformation, establishes a standardized health status quantitative index system, and realizes the interpretable mapping of comprehensive characteristics to specific health dimensions; The degradation analysis unit tracks the evolution trends of physiological, cognitive, and motor status indices based on a predefined aging trajectory model. It uses time series pattern recognition technology to capture the degradation patterns of each dimension and establish a dynamic trajectory model of the evolution of health status over time. The risk calculation unit continuously monitors the short-term change rate of exercise ability indicators. When it detects that the acceleration decreases continuously beyond the preset warning value, it automatically triggers the fall risk prediction mechanism, generates a risk window period warning signal, and associates it with the body movement fluctuation characteristics for cross-validation.

[0031] The degradation analysis unit includes: Trend extraction: using convolutional neural networks to analyze the temporal variation patterns of various health indices and extract characteristic fragments of degradation in physiological, cognitive, and motor dimensions; Aging trajectory prediction, based on the extracted degradation features, predicts the health status evolution curve in the next 90 days through a linear mixed effects model.

[0032] Specifically, trend extraction uses convolutional neural networks to automatically identify key change patterns in health index time series data, extract characteristic fragments that represent physiological degeneration, cognitive decline, and decreased motor function, capture local mutations and trend inflection points in the evolution of health status, and provide refined input features for aging trajectory modeling. Aging trajectory prediction is based on a linear mixed-effects model that integrates group degeneration patterns with individual-specific parameters. Fixed effects are used to characterize the basic age-related decline path, and random effects are used to calibrate individual health offsets to generate a personalized evolution curve of future health status, thereby achieving quantitative prediction of health risks within 90 days.

[0033] The retrospective assessment report module includes: The timeline construction unit uses the risk time trajectory and integrates health indicators to establish a unified time base for the evolution of health status. The causal chain analysis unit analyzes the causal relationship between health events based on the medical causal graph and calculates the contribution of each event to the current health risk; The visual reporting engine integrates the risk time trajectory with the causal analysis results to generate a dynamic assessment report that includes a timeline heat map, event correlation network diagram, and risk prediction curve.

[0034] Specifically, the timeline construction unit integrates discrete time point monitoring data into a continuous health status evolution sequence, associates risk trajectories with health indicators under a unified time base, establishes a traceable historical state change map, and achieves precise positioning of key nodes and panoramic visualization of state evolution; The causal chain analysis unit calculates the causal effect strength between health events based on predefined medical pathology association rules, quantifies the contribution weight of each event to the current risk status, identifies the main pathogenic factors and their transmission pathways, and forms a verifiable causal logic chain; The visual reporting engine integrates the risk timeline and causal analysis results in the form of a heat map, combines it with the correlation network diagram to reveal the interaction effects of multiple events, uses the prediction curve to depict future risk trends, and outputs an interactive dynamic assessment report with time series correlation and causal traceability.

[0035] The causal chain analysis unit includes: Causal graph construction is based on the medical causal graph, which generates a causal impact network diagram with the current health event as the root node; Contribution quantification: The contribution of each health event to the current risk is calculated using the following formula:

[0036] in, is the preset weight of the disease path in the causal graph, The risk change rate provided to the health state evolution modeling module, is the time interval from the event to the present, is a critical event indicator, i.e., 1 if it is a critical pathological stage, and 0 otherwise; Causal chain generation is to screen event nodes with a contribution greater than 0.15 and construct a multi-level causal chain with weight identification.

[0037] Specifically, the causal graph construction, based on the pathological association rules of the medical causal graph, automatically generates a multi-level causal influence network with the current health event as the tracing starting point, constructs an event-driven tree topology structure, and realizes the visual tracing of the development path of health problems; Contribution quantification is achieved by integrating path preset weights, real-time risk change rate and time attenuation factor, combined with key pathological stage identifiers, to calculate the quantitative contribution value of each event to the current health risk and establish an objective causal factor ranking mechanism.

[0038] Causal chain generation: based on the preset contribution threshold, the core influencing factor nodes are screened, a multi-level causal chain with weight identification is constructed, and a simplified causal network with clear trunk paths and prominent key nodes is automatically generated.

[0039] The decision output module includes: The medical report distribution unit sends the assessment report to the medical terminal. The report contains the causal chain of health events and the time trajectory of risks; The family warning distribution unit extracts key risk information from the assessment report, generates simplified warning notifications and pushes them to family devices.

[0040] Specifically, the medical report distribution unit encapsulates the complete causal chain of health events and risk time trajectory in the dynamic assessment report into a standard medical data format and transmits it to the medical terminal through a secure communication protocol, providing professional reports with multi-dimensional analysis to support clinical decision-making; The family warning distribution unit reconstructs the key risk information in the report that requires urgent intervention into life-oriented warning instructions, and pushes them to family user devices through the mobile communication interface to achieve immediate response to family care risks.

[0041] The Family Alert Distribution Unit includes: When the risk level exceeds the threshold of health indicators, a voice announcement will be made; Adjust the voice broadcast volume according to the user's age; Display the risk time trajectory graph on the notification interface.

[0042] Specifically, when a health risk is detected exceeding the safety threshold, the voice alarm system is triggered. The volume adaptive adjustment mechanism is driven by the user's age parameters to ensure audibility and clarity. At the same time, a risk trajectory change view is loaded on the notification interface, and a multi-dimensional perception channel is established to collaboratively output warning information, realizing three-dimensional communication of key risk situations and immediate response guidance.

[0043] Please see the attached Figure 2 , a data processing method for evaluating the health status of the elderly based on multimodal data, the method comprises the following steps: S1. Obtain electronic medical record data and medical examination reports through the FHIR interface and extract structured health indicators; S2, using image, text and image sandwich architecture to fuse monitoring data and medical text, generating multi-dimensional health feature vectors under the constraints of medical causal graph; S3, mapping the health feature vector into three-dimensional state indicators of physiological function, cognitive level and motor ability, and calculating the risk time trajectory; S4. Analyze the causal chain of historical health events and generate a dynamic assessment report based on the risk time trajectory; S5. Distribute the evaluation results to medical terminals and family members’ devices.

[0044] Specifically, an analysis foundation is established through standardized extraction of medical data, and an innovative sandwich architecture is used to fuse multimodal features under the constraints of pathological rules to generate high-dimensional vectors, which are mapped into three-dimensional quantitative health indicators and predict risk trajectories. A dynamic assessment model is constructed in combination with a temporal causal backtracking mechanism, and finally a dual-track distribution strategy is used to achieve professional reporting and simplified early warning, forming an end-to-end technical closed loop of multi-source medical data collection, fusion, modeling, evaluation and decision-making, providing dynamic assessment support for elderly health management that is both clinically interpretable and timely, significantly improving the accuracy of risk prediction and the effectiveness of intervention measures.

[0045] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A data processing system for assessing the health status of the elderly based on multimodal data, characterized in that: include: Medical data integration module, which obtains electronic medical record data and medical examination reports through the FHIR interface and extracts structured health indicators; A cross-modal causal fusion processing module, which fuses monitoring data and medical text through image, text, and image sandwich architecture to generate multi-dimensional health feature vectors under the constraints of medical causal graphs; The health status evolution modeling module maps health feature vectors into three-dimensional status indicators of physiological function, cognitive level, and motor ability, and calculates the risk time trajectory; Retrospective assessment report module, which analyzes the causal chain of historical health events and generates dynamic assessment reports based on the risk time trajectory; The decision output module is used to distribute the evaluation results to medical terminals and family devices.

2. The data processing system for elderly health status assessment based on multimodal data according to claim 1 is characterized in that: The cross-modal causal fusion processing module includes: an image feature extraction unit configured to process the medical image data through a three-dimensional convolutional network and output an anatomical structure feature matrix; A text encoding unit, configured to encode medical record text into semantic vectors using the BioBERT model; The causal constraint fusion unit divides the image features into front and back sub-matrices and embeds them into text vectors to form a sandwich structure. It applies the disease development path constraint feature weight distribution in the medical causal graph and outputs a multi-dimensional health feature vector.

3. The data processing system for elderly health status assessment based on multimodal data according to claim 2 is characterized in that: The disease development path constraints in the causal constraint fusion unit include: Path feature mapping, extracting the disease development path corresponding to the current diagnosis from the medical causal graph , the path node Encoded as a path feature vector ,in, It is a 300-dimensional embedding vector based on node types, including symptoms, diseases, and complications. The data source is the medical knowledge graph mapped by the ICD-11 code library; Attention constraint mechanism, calculating text feature vector and The cosine similarity is used as the weight constraint factor: ; in, The 768-dimensional medical record text vector output by BioBERT, is the dynamic weight factor; Weighted fusion, reorganized feature matrix: in, is the first 1 / 3 slice of the image feature, The last 2 / 3 slice of the image feature. is the Hadamard product, and the constraint is enforced when End-stage heart failure .

4. The data processing system for elderly health status assessment based on multimodal data according to claim 1 is characterized in that: The health status evolution modeling module includes: A state mapping unit converts the multidimensional health feature vector into a physiological state index, a cognitive state index, and a motor ability index through linear transformation; A degradation analysis unit analyzes the changing trends of various state indices using an aging trajectory model that includes degradation patterns in three dimensions: physiological, cognitive, and motor. The risk calculation unit monitors the rate of decline of the exercise capacity index. When the continuous monitoring shows that the exercise capacity index accelerates and drops beyond the preset threshold, it generates a fall risk time window warning and establishes an association with the extracted recent fluctuating body movement data.

5. The data processing system for elderly health status assessment based on multimodal data according to claim 4 is characterized in that: The degradation analysis unit includes: Trend extraction: using convolutional neural networks to analyze the temporal variation patterns of various health indices and extract characteristic fragments of degradation in physiological, cognitive, and motor dimensions; Aging trajectory prediction, based on the extracted degradation features, predicts the health status evolution curve in the next 90 days through a linear mixed effects model.

6. The data processing system for elderly health status assessment based on multimodal data according to claim 1 is characterized in that: The retrospective assessment report module includes: The timeline construction unit uses the risk time trajectory and integrates health indicators to establish a unified time base for the evolution of health status. The causal chain analysis unit analyzes the causal relationship between health events based on the medical causal graph and calculates the contribution of each event to the current health risk; The visual reporting engine integrates the risk time trajectory with the causal analysis results to generate a dynamic assessment report that includes a timeline heat map, event correlation network diagram, and risk prediction curve.

7. The data processing system for elderly health status assessment based on multimodal data according to claim 6 is characterized in that: The causal chain analysis unit includes: Causal graph construction is based on the medical causal graph, which generates a causal impact network diagram with the current health event as the root node; Contribution quantification: The contribution of each health event to the current risk is calculated using the following formula: in, is the preset weight of the disease path in the causal graph, The risk change rate provided to the health state evolution modeling module, is the time interval from the event to the present, is a critical event indicator, i.e., 1 if it is a critical pathological stage, and 0 otherwise; Causal chain generation is to screen event nodes with a contribution greater than 0.15 and construct a multi-level causal chain with weight identification.

8. The data processing system for elderly health status assessment based on multimodal data according to claim 1 is characterized in that: The decision output module includes: A medical report distribution unit sends an assessment report to a medical terminal, wherein the report includes a causal chain of health events and a risk time trajectory; The family warning distribution unit extracts key risk information from the assessment report, generates simplified warning notifications and pushes them to family devices.

9. The data processing system for elderly health status assessment based on multimodal data according to claim 8 is characterized in that: The family warning distribution unit includes: When the risk level exceeds the threshold of health indicators, a voice announcement will be made; Adjust the voice broadcast volume according to the user's age; Display the risk time trajectory graph on the notification interface.

10. A data processing method for assessing the health status of the elderly based on multimodal data, characterized in that: The data processing system for evaluating the health status of the elderly based on multimodal data according to any one of claims 1 to 9 comprises the following steps: S1. Obtain electronic medical record data and medical examination reports through the FHIR interface and extract structured health indicators; S2, using image, text and image sandwich architecture to fuse monitoring data and medical text, generating multi-dimensional health feature vectors under the constraints of medical causal graph; S3, mapping the health feature vector into three-dimensional state indicators of physiological function, cognitive level and motor ability, and calculating the risk time trajectory; S4. Analyze the causal chain of historical health events and generate a dynamic assessment report based on the risk time trajectory; S5. Distribute the evaluation results to medical terminals and family members’ devices.

Images