An intelligent system based on the integrated management of coronary heart disease and oral health

Abstract

This invention discloses an intelligent system for the integrated management of coronary heart disease and oral health, belonging to the interdisciplinary fields of digital healthcare and chronic disease management. The key technical points are: the system includes modules for objective data collection, data quality control and cleaning, dynamic linkage analysis, plan generation, multi-terminal intervention guidance, and multi-dimensional effect evaluation. By integrating smart hardware, AI cameras, and medical scales, it jointly extracts the pathological, oral frailty, and psychosocial objective characteristics of coronary heart disease patients; constructs a multi-dimensional coupled correlation model to dynamically predict the association risk between acute cardiovascular events and local oral infections; and achieves closed-loop intervention through real-time guidance and multi-terminal collaborative mechanisms. This solves the problem of separate management of coronary heart disease and oral health in existing health management systems, realizing cross-system collaborative management, and achieving a leap from subjective education to objective digital therapy. It significantly improves the quality of self-management for elderly coronary heart disease patients in a home environment and effectively reduces the risk of adverse health outcomes.

Description

Technical Field

[0001] This invention relates to the field of digital healthcare and chronic disease management, and more specifically, to an intelligent system based on the integrated management of coronary heart disease and oral health. Background Technology

[0002] Coronary atherosclerotic heart disease (CAD) is a major public health challenge worldwide, and its prevention and treatment have always been a core task in the medical field. With the accelerating aging of my country's society, the health management of elderly CAD patients faces increasingly severe challenges. In recent years, a growing body of medical research has revealed a close two-way correlation between oral health, especially chronic infections such as periodontitis, and the occurrence, development, and prognosis of CAD. On the one hand, the oral cavity serves as a reservoir for pathogenic microorganisms. The low-grade systemic inflammation caused by chronic infections such as periodontitis allows oral pathogens and their toxic products to breach local barriers and enter the bloodstream. This not only directly activates systemic inflammatory responses and increases the levels of key inflammatory markers such as C-reactive protein, but also continuously damages vascular endothelial function, accelerating the process of atherosclerosis. Even periodontal pathogens have been detected within atherosclerotic plaques, suggesting that these microorganisms may directly participate in and exacerbate local coronary artery lesions. Domestic research shows that the number of missing teeth is positively correlated with the risk and severity of coronary heart disease; the more teeth missing, the greater the risk. Conversely, maintaining good oral hygiene habits, such as brushing teeth effectively twice or more a day and receiving regular professional cleaning, has been proven to significantly reduce the probability of future cardiovascular events.

[0003] On the other hand, the multiple medications used in the treatment of coronary heart disease also have a significant negative impact on oral health. Many patients with coronary heart disease also have chronic conditions such as hypertension, often requiring long-term use of calcium channel blockers. A common side effect of these blockers is drug-induced gingival hyperplasia, which not only affects aesthetics and chewing function but also creates a complex environment conducive to plaque buildup, forming a vicious cycle. Simultaneously, the use of antiplatelet or anticoagulant drugs increases the risk of gingival bleeding when brushing or cleaning the mouth. Furthermore, the combined use of multiple medications can lead to dry mouth and reduced saliva production, severely weakening the mouth's natural self-cleaning and defense mechanisms, making elderly patients more susceptible to dental caries and oral fungal infections.

[0004] Against this backdrop, the emergence of the concept of "oral frailty" provides a crucial new perspective for understanding the comprehensive health status of elderly patients with coronary heart disease. Oral frailty is defined as a collection of age-related progressive declines and adverse conditions in oral function. It encompasses not only simple dental problems but also difficulties in chewing, swallowing, dry mouth, denture use issues, and the accompanying decrease in social participation and deterioration of physical and mental health. Data shows a high prevalence of oral frailty among elderly hospitalized patients. It is not an isolated local symptom but rather a manifestation and prelude to systemic frailty in the oral field. Intertwined with malnutrition, increased fall risk, and even cognitive decline, it collectively affects the overall health reserves and quality of life of the elderly. For patients with coronary heart disease, the presence of oral frailty is like adding insult to injury. Its management is not only related to basic oral function and dignity but also becomes a crucial strategic element in controlling systemic cardiovascular disease risks and improving long-term prognosis.

[0005] However, in current chronic disease management practices, elderly patients with coronary heart disease generally have weak oral health self-management abilities, and existing management models have significant structural defects and technical shortcomings:

[0006] Data monitoring is isolated and lacks objectivity: Most existing health management systems rely on patients manually entering questionnaire data, which not only suffers from severe data lag but also makes it difficult to guarantee data quality due to subjective bias and recall bias. For patients with coronary heart disease, there is a lack of routine and objective monitoring methods for their oral health status, especially functional indicators related to "oral weakness" (such as chewing and swallowing ability), resulting in insufficient basis for clinical decision-making. At the same time, there is a lack of effective anti-counterfeiting and quality control mechanisms, making it difficult to trace and verify the implementation of patients' behaviors in a home environment.

[0007] Fragmented and unpredictable risk analysis: Cardiovascular indicators and oral health data are managed in different systems or paper records, lacking a collaborative analysis model to analyze their mutual influence. This data fragmentation makes it impossible to comprehensively assess a patient's overall risk, let alone achieve dynamic causal path tracking and risk prediction for acute cardiovascular events or oral infections. Existing evaluation indicators mostly use static arithmetic scores, whose clinical predictive efficacy and individualized guidance value are extremely limited.

[0008] Intervention programs are vague and lack adherence guarantees: From a medical service perspective, there is a lack of collaboration between cardiology and dentistry. Cardiology healthcare professionals focus more on cardiac function and medication management, neglecting oral health assessments, while dentists also show insufficient consideration of specific medication contraindications and treatment safety risks for patients with coronary heart disease. From an intervention method perspective, traditional health education is mostly one-way, universal knowledge dissemination, failing to consider the dynamic changes in patients' oral function, medication use, and physical characteristics (such as limited cardiac function), and lacking systematic intervention programs based on scientific theories (such as the IMB theory). More importantly, existing intervention methods lack the ability to objectively monitor the quality of patient behavior (such as whether brushing is thorough and whether training is standardized), and do not consider safety warning mechanisms for the potential for sudden angina, arrhythmia, and other cardiovascular abnormalities in patients with coronary heart disease during home oral care, making it difficult to guarantee both intervention effectiveness and patient safety.

[0009] In summary, current health management for elderly patients with coronary heart disease faces multiple challenges, including the separation of oral and cardiovascular health management, outdated data monitoring methods, a lack of risk analysis capabilities, and a lack of scientific rigor and specificity in intervention programs. There is an urgent market demand for an intelligent management system that can achieve objective, continuous, and interconnected monitoring, integrate multimodal data (objective hardware, AI vision, and clinical indicators), possess dynamic risk prediction capabilities, and provide personalized, executable, and safety-based early warning mechanisms. This system would break down disciplinary barriers, truly achieve collaborative management of coronary heart disease and oral health, reduce the risk of adverse health outcomes, and improve the quality of life for elderly patients. Summary of the Invention

[0010] The purpose of this invention is to provide an intelligent system based on the linkage management of coronary heart disease and oral health. Through systematic information intervention, motivation stimulation and behavioral skills training, it can improve the oral health self-management ability of elderly patients with coronary heart disease, thereby indirectly improving their cardiovascular health outcomes.

[0011] The above-mentioned technical objective of this invention is achieved through the following technical solution: an intelligent system based on the coordinated management of coronary heart disease and oral health, the intelligent system comprising:

[0012] 1. Objective data acquisition module

[0013] As the system's data input layer, a multi-source heterogeneous data acquisition mechanism is adopted. Data is collected jointly through smart IoT hardware, mobile cameras, and third-party health platform interfaces.

[0014] Clinical biochemical and time-series data of coronary heart disease: heart rate and blood pressure fluctuation curves continuously collected by smart wearable devices; medication information for coronary heart disease and core indicators of systemic inflammation.

[0015] Objective oral health and visual data: dynamic data transmitted by the smart electric toothbrush; intraoral and gingival imaging data collected via the front-facing camera; acoustic data of chewing and swallowing.

[0016] Social and psychological assessment data: scale assessments and frequency of in-app interactions between family members / caregivers.

[0017] 2. Data quality control and cleaning module

[0018] Used for real-time quality assessment and cross-validation of the collected data:

[0019] Real-time quality control of visual data: When a patient's oral cavity is being photographed, the system evaluates the preview frames of the oral cavity images in real time, calculating image sharpness, illumination, and angle. If the thresholds are not met, the system intercepts the image and prompts for adjustment in real time.

[0020] Subjective-objective cross-validation: The timestamps of the data subjectively reported by patients are compared with the data returned by smart hardware to intercept false attendance records.

[0021] 3. Dynamic Linkage Analysis Module

[0022] Joint modeling of multidimensional coupled features based on graph neural network-causal factorization structure:

[0023] A multidimensional heterogeneous atlas was constructed, including a map of oral function fluctuations, a map of physiological metabolic changes, a map of drug use characteristics, and a map of social interaction effects.

[0024] We extracted the interaction causal factors between short-term oral inflammation and long-term cardiovascular outcomes, and output dynamic risk prediction scores for acute cardiovascular events and local oral infections.

[0025] 4. Solution Generation Module

[0026] Based on the results of linkage analysis and the Information-Motivation-Behavior Skills (IMB) theory, a personalized digital therapy intervention plan is generated, which includes medication guidance, oral function rehabilitation, psychological intervention and social support mobilization.

[0027] 5. Intervention Implementation and Guidance Module

[0028] The solution is implemented through multi-device collaboration. The built-in augmented reality rendering module overlays a virtual dentist character onto real-time camera footage, guiding users to adjust the shooting angle, distance, and lighting conditions. Simultaneously, a multi-modal safety warning mechanism is included. If an abnormal heart rate change is detected during brushing or severe gum bleeding is identified by AI, the task is immediately interrupted and an alarm is triggered.

[0029] 6. Multidimensional Effect Evaluation Module

[0030] The model output is posteriorly calibrated based on Monte Carlo and confidence interval adjustment mechanisms. Through dynamic trajectory tracking, objective behavioral compliance rate, oral function improvement rate, and cardiovascular sign stability are monitored to form an adaptive and optimized management closed loop.

[0031] In summary, the present invention has the following beneficial effects:

[0032] 1. Shift from passive recording to proactive monitoring to improve data objectivity: By integrating IoT hardware such as smart toothbrushes and wearable devices, the system can continuously and objectively collect key behavioral and physiological data without increasing the burden on patients. This solves the pain points of traditional systems, which rely on a single data source and manual input, and provides a reliable foundation for accurate analysis.

[0033] 2. Achieve linked analysis and early warning of oral and cardiovascular risks: Instead of looking at oral or cardiac problems in isolation, deep learning models are used to quantify the dynamic impact between the two, accurately identify cardiovascular risks caused by oral problems, provide doctors and patients with more forward-looking early warning information, and meet the urgent clinical need for comprehensive risk assessment.

[0034] 3. The intervention program is practical and adherent: By introducing augmented reality (AR) technology for behavioral guidance, abstract health education is transformed into intuitive and standardized operational instructions, greatly improving the learning effectiveness and adherence of patients (especially elderly patients). This makes the system not only a management tool, but also a "personal rehabilitation coach".

[0035] 4. Forming a self-optimizing closed-loop management system: The dynamic effect evaluation module ensures that intervention strategies can be continuously iterated based on changes in the patient's condition and feedback, keeping management services constantly updated rather than being a one-time fix. This "gets smarter with use" characteristic is the core of the product's long-term market competitiveness.

[0036] 5. Clear Product Conversion Path and Market Demand: This system is directly targeted at cardiac rehabilitation centers, dental clinics, community health service centers, elderly care institutions, and the families of elderly patients with coronary heart disease. Its modular design supports multiple profit models, including hardware sales, software service subscriptions, and data analysis reports, demonstrating clear product conversion value and genuine purchasing demand. Detailed Implementation

[0037] The present invention will now be described in further detail.

[0038] Example: An intelligent system based on the integrated management of coronary heart disease and oral health.

[0039] The system provided by the present invention will be further described below with reference to specific examples.

[0040] A typical patient, Mr. Zhao, 75 years old, underwent coronary stent implantation, and has been taking dual antiplatelet therapy long-term. He lives alone and has early-stage oral weakness. The single-run closed-loop process of this system is as follows:

[0041] 1. Multi-source objective data acquisition and AI visual quality control

[0042] At 8 a.m., Mr. Zhao prepared for oral care. He picked up his smart electric toothbrush, which was connected to the Internet of Things, and opened the mobile app to enter augmented reality interaction mode.

[0043] Data Acquisition and Quality Control: The system first detects that the phone is held upside down and automatically adjusts the interface. Guided by the virtual dentist, Mr. Zhao opens his mouth, and the system extracts visual features in real time using a multi-stage cascaded convolutional neural network. The system detects that the lighting is dim and the angle is off when photographing Mr. Zhao's left posterior teeth area, and the real-time evaluation indicators do not meet the threshold conditions. The virtual character immediately issues a voice message and instructs him with an arrow to "adjust slightly to the left by 15 degrees and move closer to the light source." After the adjustment is successful, the system automatically captures a high-definition still image. Simultaneously, the smart toothbrush transmits the brushing pressure value (indicating excessive force), the smart bracelet transmits real-time heart rate and blood pressure data, and the phone's microphone records his subsequent breakfast chewing audio (identifying a chewing rate significantly lower than the age-appropriate baseline, suggesting oral health issues).

[0044] 2. Cross-validation and dynamic linkage analysis

[0045] Anomaly interception and verification: The system backend retrieved Zhao's electronic medical record data and confirmed that he was in a period of high adherence to dual antiplatelet therapy.

[0046] Dynamic Linkage Analysis Module: The model inputs the aforementioned objective data, performing multidimensional coupled analysis on "antiplatelet drug use (prone to bleeding)," "severely red and congested gingival plaques identified by AI vision," "high-pressure, forceful brushing behavior detected by the smart toothbrush," and "weakness characteristics of chewing." The model outputs a high-risk warning: The probability of Mr. Zhao experiencing severe periodontal bleeding is extremely high, and the accompanying local infection entering the bloodstream and causing systemic inflammation, thereby inducing an acute cardiovascular event, has exceeded the upper bound of the safety confidence level.

[0047] 3. Real-time intervention and guidance, and multi-terminal collaboration

[0048] Real-time intervention on the patient's end: The moment the warning was triggered, the system forcibly reduced the vibration frequency of the smart electric toothbrush via Bluetooth, switching it to a gentle gum-care mode. At the same time, the virtual dentist on the phone screen switched from regular guidance to a red warning state, gently but seriously reminding Mr. Zhao to stop brushing his teeth aggressively and demonstrating the correct Bass brushing technique with gentle movements.

[0049] Collaborative intervention between medical staff and family: The high-risk warning and key gingival congestion image slices were simultaneously pushed to the attending nurse's management platform. The system automatically set Zhao's "intervention priority" to the highest level. The system sent a structured task instruction to his daughter's "family collaboration terminal": "Father's gingival bleeding risk is extremely high today and he has been depressed recently. Please initiate a video call within the system at 8 pm tonight to supervise and encourage father to complete gentle oral care through the AR co-view function."

[0050] 4. Dynamic effect tracking and closed-loop optimization

[0051] Two weeks later, the multidimensional effect assessment module, through time-series comparison, found that Zhao's brushing pressure had stabilized within the safe threshold, AI visual continuous monitoring showed that the area of ​​gingival redness and swelling had decreased by 40%, and the chewing audio rate had increased. Combined with the increased frequency of interaction with family members, his social deterioration index decreased. The system determined that the intervention at this stage had achieved its goals and automatically used Monte Carlo sampling to generate a new and more stable confidence interval for cardiovascular risk prediction for the next cycle, completing a complete closed loop from monitoring, assessment, early warning to precise intervention.

[0052] This specific embodiment is merely an explanation of the present invention and is not intended to limit the invention. After reading this specification, those skilled in the art can make modifications to this embodiment without contributing any inventive step, but such modifications are protected by patent law as long as they are within the scope of the claims of the present invention.

Claims

1. An intelligent system based on the integrated management of coronary heart disease and oral health, characterized by: The intelligent system includes: The objective data acquisition module, as the data input layer of the system, jointly collects objective vital signs time-series data, AI visual image data, clinical biochemical data, and social and psychological status data of elderly patients with coronary heart disease through smart IoT hardware, mobile camera, and hospital electronic medical record interface. The data quality control and cleaning module is used to perform real-time quality assessment and cross-validation on the collected multi-source heterogeneous data, intercept unqualified visual images, and perform outlier filtering and interpolation repair on hardware time-series data to ensure the objectivity and validity of the input data. The dynamic linkage analysis module constructs a multidimensional coupled correlation model of oral frailty, cardiovascular risk, and social psychological characteristics based on graph neural networks. It analyzes the causal path between real-time vital signs and long-term behavior, and outputs dynamic risk prediction scores for acute cardiovascular events and local oral infections. The treatment plan generation module, based on the results of the linkage analysis and the information-motivation-behavior skills theory, generates personalized digital therapy intervention plans that cover medication guidance, oral function rehabilitation, psychological intervention, and social support mobilization. The intervention execution and guidance module implements the plan through multi-terminal collaboration, which specifically refers to the collaboration between the patient, medical staff, and family members. It also uses real-time visual and acoustic interaction to guide patients to complete standard oral care and monitoring tasks. The multidimensional efficacy assessment module dynamically monitors patients' improvements in compliance, reversal of oral weakness, and stability of cardiovascular signs by comparing historical time-series data with post-intervention data, forming an adaptive and optimized management loop.

2. The intelligent system based on the integrated management of coronary heart disease and oral health according to claim 1, characterized in that: The data collected by the objective data acquisition module specifically includes: Clinical biochemical and time-series data of coronary heart disease: heart rate, blood pressure fluctuation curves and blood oxygen saturation continuously collected through smart wearable devices; coronary heart disease course and medication information obtained through system interface, the medication information including anticoagulant / antiplatelet drugs that affect the oral environment and calcium channel blockers and systemic inflammatory core indicators including high-sensitivity C-reactive protein; Objective oral health and visual data: behavioral dynamics data automatically transmitted by the smart electric toothbrush, including daily brushing time, force, and blind spots; intraoral and gingival imaging data collected periodically by the patient's app camera; and acoustic data of chewing and swallowing collected by the microphone. Social and psychological assessment data: including basic information such as living alone status and average monthly household income per capita, as well as scale-based oral health-related self-efficacy data, social weakness index, and frequency of in-app interactions with family members / caregivers.

3. The intelligent system based on the integrated management of coronary heart disease and oral health according to claim 1, characterized in that: The control mechanism of the data quality control and cleaning module includes: Real-time quality control of visual data: When patients use mobile devices to take oral images, the system calculates the image clarity, illumination uniformity, and whether the shooting angle covers key areas in real time. If the preset threshold conditions are not met, the system will intercept and trigger an interface prompt to guide the user to adjust their posture and retake the image. Cross-validation mechanism for subjective and objective data: The patient's subjective compliance data filled in on the APP is compared with the objective data returned by the smart hardware by timestamp comparison. Mismatched data is downgraded and a verification reminder is triggered on the family collaboration terminal.

4. The intelligent system based on the integrated management of coronary heart disease and oral health according to claim 1, characterized in that: The construction method of the dynamic linkage analysis module is as follows: The collected standardized sequences were constructed into a multidimensional heterogeneous atlas including oral function fluctuation map, physiological metabolism change map, medication characteristic map and social interaction influence map. Graph convolutional neural network was used to model and extract the interaction causal factors between short-term acute oral inflammation and long-term cardiovascular outcomes in patients. By combining historical case follow-up outcomes, the probability of an individual experiencing adverse coronary events in a specific future period is output, and key nodes leading to high risk are identified.

5. The intelligent system based on the coordinated management of coronary heart disease and oral health according to claim 1, characterized in that: The intervention execution and guidance module enables the implementation of the intervention plan from the patient's end, medical staff's end, and family collaboration end. Its multi-terminal collaboration and interaction mechanism is as follows: Patient-side interaction and safety alerts: Includes a real-time operation guidance subunit based on computer vision or augmented reality; during oral care, the system corrects the patient's movement angle and coverage blind spots in real time by overlaying virtual instructions or voice on the screen; at the same time, the terminal has a built-in multimodal safety alert mechanism. If the smart wearable device detects abnormal heart rate changes or the AI ​​vision identifies severe gum bleeding during brushing teeth or eating check-in, the system will immediately interrupt the current task and trigger an emergency alarm to medical staff and family members. Family-led intervention: In addition to receiving synchronous reminders about the patient's task completion status, the system automatically pushes motivational messages and collaborative supervision tasks to family members based on the patient's recent social weakness index and check-in enthusiasm, quantitatively assessing and strengthening family social support. Dynamic monitoring by healthcare professionals: Automatically aggregates high-risk early warning information and objective compliance reports; healthcare professionals can view abnormal visual slices and abnormal physiological time-series fluctuation graphs highlighted by the system, and issue adjusted intervention plans with one click.

6. The intelligent system based on the integrated management of coronary heart disease and oral health according to claim 1, characterized in that: The multi-dimensional performance evaluation module employs a dynamic trajectory tracking and adaptive closed-loop optimization mechanism, the specific process of which is as follows: Multidimensional dynamic index quantification: abandoning static questionnaire scores, outputting three-dimensional evaluation indicators in the form of time series: (1) Objective behavioral compliance rate, specifically the frequency / duration of smart toothbrush activation reaching the target rate; (2) Oral function and signs improvement rate, specifically the plaque area reduction rate, gingival redness and swelling reduction rate and chewing audio rate improvement value calculated based on continuous visual images; (3) Cardiovascular signs stability, specifically the improvement value of heart rate variability and the proportion of blood pressure reaching the target stable period; Effect evaluation based on confidence intervals: By extracting the range and skewness coefficients of the above time series indicators, and combining them with Monte Carlo sampling, a predicted confidence interval for the improvement effect is constructed; if the actual improvement curve falls below the preset lower confidence bound, the intervention at this stage is deemed unsatisfactory. Adaptive closed-loop optimization and attribution: When the intervention is deemed unsatisfactory, the system automatically initiates attribution analysis to locate the mismatch characteristics; The system automatically adjusts the parameters in the plan generation module based on the attribution results and seamlessly pushes the new plan to the intervention execution module, forming a digital therapy closed loop of objective monitoring, dynamic evaluation, automatic attribution, and strategy iteration.