CN-122000052-A - Intelligent body measurement health management method, system and body measurement instrument integrating AI large model

CN122000052ACN 122000052 ACN122000052 ACN 122000052ACN-122000052-A

Abstract

The invention relates to the technical field of health management, in particular to an intelligent body measurement health management method, system and body measurement instrument integrating an AI large model, wherein the method generates an individual physiological steady-state baseline track by collecting multi-dimensional physiological sign time sequence data and analyzing a biological rhythm; and extracting texture features of the tissue microstructure by combining continuous dynamic body measurement image data, and constructing an organ function reserve dynamic map. And (3) merging the data to establish a multi-level chronic disease risk gradient model, and outputting a multi-time scale disease activity prediction curve. And identifying high-risk target organs based on the images, and deducting personalized intervention window periods by combining with a prediction curve to form an organ-dividing and staged accurate intervention task set. By dynamically scheduling individual health intervention resources and collecting feedback data, a physiological steady-state baseline is reconstructed in real time based on treatment effect attribution and threshold drift detection, a closed-loop dynamic health management strategy is formed, and accurate early warning and individual intervention of chronic disease risk are realized.

Inventors

CHENG YUNZE
LI YINGTAO
LI YINGRI
WANG MINGDONG
Tan Jiaoli

Assignees

广东米果智能设备有限公司

Dates

Publication Date: 20260508
Application Date: 20260119

Claims (10)

1. An intelligent body measurement health management method integrating an AI large model is characterized by comprising the following steps: The method comprises the steps of collecting individual multidimensional physiological sign time sequence data, carrying out biological rhythm analysis and steady-state offset modeling on the individual multidimensional physiological sign time sequence data, and generating an individual physiological steady-state baseline track; performing tissue microstructure texture characterization extraction and cross-modal function coupling analysis based on the individual continuous dynamic body measurement image data to generate an organ function reserve dynamic map; fusing the individual physiological steady-state baseline track and the organ function reserve dynamic map, constructing a multi-level chronic disease progression risk gradient model, and outputting a multi-time scale disease activity prediction curve; Combining a multi-time scale disease activity prediction curve, performing personalized intervention window period deduction on each target organ region, and generating an organ-dividing and staged accurate intervention task set; according to the accurate intervention task set, implementing dynamic adaptive scheduling under space-time constraint on the individualized healthy intervention resources, and synchronously collecting intervention response feedback data; And performing curative effect attribution analysis and threshold drift detection based on intervention response feedback data, and driving the physiological steady-state baseline track to reconstruct in real time to form a closed-loop dynamic threshold health management strategy.
2. The method for intelligent physical testing health management of an integrated AI large model of claim 1, wherein said collecting individual multidimensional physiological sign time series data, performing biorhythm analysis and steady state shift modeling on said individual multidimensional physiological sign time series data, generating an individual physiological steady state baseline trajectory comprises: The method comprises the steps of cooperatively collecting individual multidimensional physiological sign time sequence data through a wearable sensing array and a sensorless sensing terminal, wherein the individual multidimensional physiological sign time sequence data at least comprises heart rate variability, skin electric response, respiratory tidal volume, microcirculation blood oxygen saturation and sublingual microvascular flow rate; Performing circadian rhythm phase decoupling and autonomic nerve tension spectrum decomposition on the individual multidimensional physiological sign time sequence data to generate an individual biological rhythm parameter matrix; constructing a physiological steady state kinetic equation based on the individualized biological rhythm parameter matrix, and solving a steady state attractor topological structure; Carrying out long-term steady-state offset quantization by using a steady-state attractor topological structure to generate a steady-state offset entropy value sequence; fusing the steady-state offset entropy value sequence and the individualized biological rhythm parameter matrix, and training an individualized physiological steady-state baseline generator; outputting an individualized physiological steady-state baseline track through the individualized physiological steady-state baseline generator, wherein the individualized physiological steady-state baseline track comprises a dynamic confidence interval and a pathological disturbance sensitive zone.
3. The method for intelligent body measurement health management of an integrated AI large model of claim 1, wherein said obtaining individual continuous dynamic body measurement image data, performing tissue microstructure texture characterization extraction and cross-modal functional coupling analysis based on said individual continuous dynamic body measurement image data, generating an organ functional reserve dynamic map, comprises: synchronously acquiring individual continuous dynamic body measurement image data through a multispectral skin-mucosa imager and tongue microstructure optical coherence tomography; extracting the multi-scale vascular network skeleton of the individual continuous dynamic body measurement image data to construct a microcirculation topological connection diagram; Fusing the microcirculation topological connection diagram and the infrared thermal metabolism image, and performing tissue perfusion-metabolism coupling modeling to generate an organ-level functional reserve dynamic map; Calculating the functional redundancy index and the compensatory attenuation slope of each target organ based on the organ-level functional reserve dynamic map; performing cross-organ correlation analysis on the functional redundancy index and the compensation attenuation slope, and identifying a functional compensation imbalance hub node; and reversely mapping the pivot node serving as an anchor point to an anatomical structure space to generate an organ function reserve dynamic map, wherein the organ function reserve dynamic map comprises a space positioning label set.
4. The method for intelligent body measurement health management of integrated AI large model of claim 3, wherein said performing multi-scale vascular network skeleton extraction on said individual continuous dynamic body measurement image data to construct a microcirculatory topological connection graph comprises: performing light scattering consistency calibration among multispectral channels on individual continuous dynamic body measurement image data; Performing texture sharpening on the calibrated image by adopting a self-adaptive local contrast enhancement algorithm to generate an enhanced texture image; constructing a multi-scale Gabor filter group based on the enhanced texture image, and extracting directional texture characteristics of the tissue microstructure; Training a lightweight U-Net segmentation model by using directional texture characteristics to realize pixel-level segmentation of the microvascular network; and performing graph theory optimization on the segmentation result, removing the pseudo-connected branches, complementing the broken vessel segments, and outputting a microcirculation topological connection graph.
5. The integrated AI large-model intelligent body-testing health management method of claim 1, wherein the fusing of individual physiological steady-state baseline trajectories and organ function reserve dynamic profiles, constructing a multi-level chronic disease progression risk gradient model, outputting a multi-time scale disease activity prediction curve, comprises: Mapping the individualized physiological steady-state baseline track to a space-time coordinate system of an organ function reserve dynamic map, and constructing a cross-modal steady-function joint characterization space; defining a multi-dimensional risk potential energy field of chronic disease progression in the cross-modal steady-function joint characterization space, wherein the potential energy field is formed by inflammatory factor fluctuation entropy, metabolite accumulation gradient and neuroendocrine axis disturbance amplitude; simulating a disease progress path by adopting a Lagrangian mechanical frame based on the multidimensional risk potential energy field to generate a disease activity evolution manifold; Inputting a pre-trained multi-modal large model, wherein the multi-modal large model fuses a clinical guideline knowledge graph, real world queue survival data and individual gene polymorphism information to carry out causal inference enhancement on disease activity evolution manifold; And outputting disease activity prediction curves covering multi-stage time granularity, wherein each disease activity prediction curve is attached with an uncertainty quantitative index and a key turning point early warning mark.
6. The method for intelligent body-building health management of an integrated AI large model as set forth in claim 1, wherein said identifying high risk target organ regions based on individual continuous dynamic body-building images, performing personalized intervention window period deduction on each target organ region in combination with a multi-time scale disease activity prediction curve, generating a split-organ, staged accurate intervention task set, comprises: Identifying four high-risk target organ areas of liver, kidney, pancreas and retina based on a space positioning label set of organ function reserve dynamic maps in the continuous dynamic body measurement images; Performing pathological image feature transfer learning on each target organ region, and identifying subclinical pathological change markers of early fibrosis, microangioma and beta cell apoptosis; Calculating intervention urgency scores of all target organs according to the detection intensity, the spatial distribution density and the deviation degree from a physiological steady-state baseline track of the subclinical lesion markers; constructing an intervention resource space-time reachability map based on individual daily activity tracks and environment exposure data, and extracting a feasible intervention path conforming to the circadian rhythm constraint; performing pareto optimal matching on the intervention urgency score and a feasible intervention path, and deducing an optimal intervention starting time window and minimum effective intervention intensity of each target organ; Integrating the target organ intervention time window, the intensity threshold and the resource constraint condition to generate an accurate intervention task set containing the execution priority, the dose gradient and the curative effect monitoring node.
7. The intelligent body measurement health management method of an integrated AI large model of claim 1, wherein said identifying callable individual health intervention resources, performing a dynamic adaptive scheduling under space-time constraints on individual health intervention resources based on said set of precise intervention tasks, and synchronously collecting intervention response feedback data, comprises: identifying callable individual health intervention resources, including nutrient slow-release microcapsules, percutaneous acupoint electrical stimulation parameter sets, personalized exercise prescription libraries and intestinal flora regulator combinations; Establishing a digital twin body for the individualized healthy intervention resource, simulating pharmacodynamic response of the digital twin body under the individual physiological steady-state baseline track, and generating a resource efficiency-toxicity balance matrix; And carrying out dynamic resource scheduling by adopting a multi-target reinforcement learning algorithm according to the resource efficiency-toxicity balance matrix and the accurate intervention task set, outputting an intervention execution instruction sequence with space-time coordinates, and synchronously acquiring intervention response feedback data.
8. The integrated AI large model intelligent body test health management method of claim 1, wherein the performing therapy effect attribution analysis and threshold drift detection based on intervention response feedback data drives physiological steady state baseline trajectory real-time reconstruction to form a closed loop dynamic threshold health management strategy comprises: carrying out multi-source heterogeneous signal alignment on intervention response feedback data, wherein the multi-source heterogeneous signal alignment comprises continuous blood sugar fluctuation spectrum, urine microalbumin/creatinine ratio dynamic change, inflammatory factor time sequence concentration and subjective symptom diary text; Constructing an intervention curative effect attribution graph neural network, and analyzing contribution degree weights of each intervention measure to functional reserves of different target organs; Triggering a threshold drift detection mechanism when the functional redundancy index of any target organ is lower than a preset safety threshold and the duration exceeds a preset duration; Based on the threshold drift detection result, retraining the individual physiological steady-state baseline generator, updating the physiological steady-state baseline track and the dynamic confidence interval, and completing the closed-loop dynamic threshold health management strategy iteration.
9. An integrated AI large model intelligent body measurement health management system, the system comprising: The data modeling module is used for collecting the time sequence data of the individual multidimensional physiological signs, analyzing the biological rhythm and modeling the steady-state offset of the time sequence data of the individual multidimensional physiological signs, and generating an individual physiological steady-state baseline track; The image analysis module is used for acquiring individual continuous dynamic body measurement image data, extracting tissue microstructure texture characterization and cross-modal function coupling analysis based on the individual continuous dynamic body measurement image data, and generating an organ function reserve dynamic map; the risk modeling module is used for fusing the individual physiological steady-state baseline track and the organ function reserve dynamic map, constructing a multi-level chronic disease progression risk gradient model and outputting a multi-time scale disease activity prediction curve; The intervention deduction module is used for identifying high-risk target organ areas based on individual continuous dynamic body measurement images, executing personalized intervention window period deduction on each target organ area by combining a multi-time scale disease activity prediction curve, and generating an organ-dividing and staged accurate intervention task set; the resource scheduling module is used for identifying callable individualized healthy intervention resources, implementing dynamic adaptive scheduling under space-time constraint on the individualized healthy intervention resources according to the accurate intervention task set, and synchronously collecting intervention response feedback data; The closed-loop management module is used for carrying out curative effect attribution analysis and threshold drift detection based on intervention response feedback data, driving the real-time reconstruction of the physiological steady-state baseline track, and forming a closed-loop dynamic threshold health management strategy.
10. A body-measuring apparatus comprising a memory, a processor and an integrated AI large model of an intelligent body-measuring health management program stored on the memory and executable on the processor, the integrated AI large model of an intelligent body-measuring health management program configured to implement the steps of the integrated AI large model of an intelligent body-measuring health management method of any of claims 1-9.

Description

Intelligent body measurement health management method, system and body measurement instrument integrating AI large model Technical Field The invention relates to the technical field of health management, in particular to an intelligent body measurement health management method, system and body measurement instrument integrating an AI large model. Background Along with the continuous rising of chronic disease prevalence, the traditional health management mode is difficult to realize accurate layering and individuation intervention of chronic disease risks due to the dependence on single-dimension physiological indexes, lack of dynamic physiological rhythm analysis and cross-modal correlation analysis of organ function reserves. In the prior art, physiological steady state assessment is mostly based on a static threshold value, time sequence characteristics of individual biological rhythm differences and steady state offset are not fully considered, body measurement image analysis is limited to qualitative observation of structures, coupled modeling of tissue microstructure textures and functional states is lacked, and a risk prediction model cannot effectively fuse physiological baseline tracks and organ functional dynamic maps, so that time scale of disease activity prediction is single, and risk gradient division is rough. In addition, the intervention strategy preparation always adopts a cut-off mode, the accurate intervention window period is dynamically deduced without combining individual characteristics and disease progress, and the scheduling of healthy intervention resources lacks a dynamic adaptation mechanism under space-time constraint, so that a closed-loop management system based on intervention response feedback is difficult to form. The foregoing is provided merely for the purpose of facilitating understanding of the technical solutions of the present invention and is not intended to represent an admission that the foregoing is prior art. Disclosure of Invention The invention mainly aims to provide an intelligent body measurement health management method, system and body measurement instrument integrating an AI large model, and aims to solve the technical problems that a traditional health management mode depends on a single-dimension physiological index, lacks dynamic circadian rhythm analysis and organ function cross-modal correlation analysis, is based on a static threshold value in steady state evaluation, is lack of fusion of a risk prediction model, lacks dynamic adaptation and closed-loop management in intervention strategy one-step and health intervention resource scheduling, and is difficult to realize accurate layering of chronic disease risk and individualized intervention. In order to achieve the above purpose, the present invention provides an intelligent body testing health management method integrating an AI large model, the method comprising the steps of: The method comprises the steps of collecting individual multidimensional physiological sign time sequence data, carrying out biological rhythm analysis and steady-state offset modeling on the individual multidimensional physiological sign time sequence data, and generating an individual physiological steady-state baseline track; performing tissue microstructure texture characterization extraction and cross-modal function coupling analysis based on the individual continuous dynamic body measurement image data to generate an organ function reserve dynamic map; fusing the individual physiological steady-state baseline track and the organ function reserve dynamic map, constructing a multi-level chronic disease progression risk gradient model, and outputting a multi-time scale disease activity prediction curve; Combining a multi-time scale disease activity prediction curve, performing personalized intervention window period deduction on each target organ region, and generating an organ-dividing and staged accurate intervention task set; according to the accurate intervention task set, implementing dynamic adaptive scheduling under space-time constraint on the individualized healthy intervention resources, and synchronously collecting intervention response feedback data; And performing curative effect attribution analysis and threshold drift detection based on intervention response feedback data, and driving the physiological steady-state baseline track to reconstruct in real time to form a closed-loop dynamic threshold health management strategy. Optionally, the acquiring the individual multidimensional physiological sign time sequence data, performing biological rhythm analysis and steady-state offset modeling on the individual multidimensional physiological sign time sequence data, and generating an individual physiological steady-state baseline track comprises the following steps: The method comprises the steps of cooperatively collecting individual multidimensional physiological sign time sequence data through a wearable sensi