CN-122000041-A - Systems and methods for noninvasive cardiovascular health assessment

Abstract

A computer-implemented method for non-invasive cardiovascular health assessment includes receiving, at a processor, a physiological signal of a subject, wherein the physiological signal includes at least a cardiac ECG signal (electrical signal) and a PPG signal (photoplethysmogram signal), processing the ECG signal and the PPG signal by an AI model (artificial intelligence model) executed by the processor, and estimating, by the AI model, one or more demographic and anthropometric parameters based on the processing of the ECG signal and the PPG signal.

Inventors

• Zhi Jieyi
• CHEN HAOMIN
• LV LIANGYI
• Lv Liangting
• GONG ZHENG

Assignees

• 香港心脑血管健康工程研究中心有限公司

Dates

Publication Date

20260508

Application Date

20251104

Priority Date

20241107

Claims (20)

1. 1. A computer-implemented method for non-invasive cardiovascular health assessment, comprising: Receiving, at a processor, a physiological signal of a subject; Wherein the physiological signals include at least a cardiac ECG signal (electrical signal) and a PPG signal (photoplethysmography signal); an AI model (artificial intelligence model) executed by the processor, processes the ECG signal and the PPG signal, and One or more demographic and anthropometric parameters are estimated by the AI model based on processing of the ECG and PPG signals.
2. 2. The method of claim 1, wherein the processing step comprises: Pre-processing the received ECG and PPG signals; feature extraction of the preprocessed ECG signal and PPG signal, and Selecting one or more optimal features, and Wherein the demographic and anthropometric parameters are estimated based on the one or more optimal characteristics.
3. 3. The method of claim 2, wherein the processing step further comprises performing temporal processing on the pre-processed ECG and PPG signals.
4. 4. The method of claim 3, wherein the temporal processing comprises capturing a temporal correlation and a spatial relationship between the one or more optimal features.
5. 5. The method of claim 1, wherein the demographic parameters include age, gender, and the one or more anthropometric features include weight, height, and Body Mass Index (BMI).
6. 6. The method as recited in claim 1, further comprising: calculating a cardiovascular age of the subject based on the estimated demographic and anthropometric parameters by the AI model, and Outputting the cardiovascular age, and/or comparing the calculated cardiovascular age to a physiological age of the subject to identify a risk of vascular aging.
7. 7. The method of claim 6, further comprising tracking changes in the estimated BMI or weight over time to detect abnormal trends and presenting the estimated changes in BMI or weight on a user interface.
8. 8. The method as recited in claim 1, further comprising: Generating visual indicators to show estimates of the demographic and anthropometric parameters, and Presenting the visual indicator on a user interface, and Wherein the visual indicators include one or more of a saliency map, a heat map, or an attention heat map.
9. 9. A method according to claim 3, wherein the AI model comprises an RNN model (recurrent neural network model), wherein the RNN model is trained to perform the step of processing the received ECG and PPG signals and to output an estimate of one or more demographic and anthropometric parameters.
10. 10. A system for non-invasive cardiovascular health monitoring, comprising: a computing device comprising a processor and a storage unit, the processor and the storage unit being operably coupled; A user interface operatively coupled to or integrated in the computing device; Wherein the storage unit is configured to store an AI model (artificial intelligence model), wherein the AI model is executable by the processor, the storage unit further comprising instructions that, when executed by the processor, cause the processor to: A physiological signal of a subject is received, Wherein the physiological signals include at least a cardiac ECG signal (electrical signal) and a PPG signal (photoplethysmography signal); processing the ECG signal and the PPG signal by the AI model; estimating one or more demographic and anthropometric parameters based on processing the ECG signal and the PPG signal by the AI model, and Displaying the estimated demographic and anthropometric parameters on the user interface.
11. 11. The system of claim 10, wherein the processor is programmed to: Pre-processing the received ECG and PPG signals; feature extraction of the preprocessed ECG signal and PPG signal, and Selecting one or more optimal features, and Wherein the demographic and anthropometric parameters are estimated based on the one or more optimal characteristics.
12. 12. The system of claim 11, wherein the processor is programmed to perform temporal processing on the pre-processed ECG and PPG signals.
13. 13. The system of claim 12, wherein during temporal processing, the processor is programmed to capture a temporal correlation and spatial relationship between the one or more optimal features.
14. 14. The system of claim 10, wherein the demographic parameters include age, gender, and the one or more anthropometric features include weight, height, and Body Mass Index (BMI).
15. 15. The system of claim 10, wherein the processor is programmed to: calculating a cardiovascular age of the subject based on the estimated demographic and anthropometric parameters by applying the AI model, and Outputting the cardiovascular age, and/or comparing the calculated cardiovascular age to a physiological age of the subject to identify a risk of vascular aging.
16. 16. The system of claim 15, wherein the processor is programmed to track changes in the estimated BMI or weight over time to detect abnormal trends and present the changes in the estimated BMI or weight on a user interface.
17. 17. The system of claim 10, wherein the processor is further programmed to: generating visual indicators to show an estimate of the demographic and anthropometric parameters; presenting the visual indicator on the user interface, and Wherein the visual indicators include one or more of a saliency map, a heat map, or an attention heat map.
18. 18. The system of claim 11, wherein the AI model comprises an RNN model (recurrent neural network model), wherein the RNN model is trained to perform the step of processing the received ECG and PPG signals and to output an estimate of one or more demographic and anthropometric parameters.
19. 19. The system of claim 18, wherein the AI model is pre-trained on a dataset of unlabeled ECG signals, PPG signals, or both using a self-supervised learning framework, wherein the self-supervised learning framework is a masked self-supervised learning framework, fragments of the unlabeled signals are masked, and the AI model is trained to reconstruct the masked fragments.
20. 20. A wearable device for non-invasive cardiovascular health monitoring, comprising: A sensor array for continuously acquiring an ECG signal and a PPG signal from a user; a computing device comprising a processor and a storage unit, the processor and the storage unit being operably coupled; The computing device is operably coupled to the sensor array; The storage unit for storing an AI model (artificial intelligence model), wherein the AI model is executable by the processor, the storage unit further comprising instructions that, when executed by the processor, cause the processor to: Receiving a physiological signal of a subject, wherein the physiological signal comprises at least a cardiac ECG signal (electrical signal) and a PPG signal (photoplethysmography signal); processing the ECG signal and the PPG signal by the AI model, and One or more demographic and anthropometric parameters are estimated by the AI model based on processing of the ECG and PPG signals.

Description

Systems and methods for noninvasive cardiovascular health assessment Technical Field The present disclosure relates to a system and method for non-invasive cardiovascular health assessment. The present disclosure may also relate to a system and method for non-invasive cardiovascular health assessment based on an ECG signal and a PPG signal. Background Cardiovascular health remains a central challenge in contemporary medicine because it directly affects life expectancy and quality of life. Accurate estimation of demographic and anthropometric characteristics such as age, sex, height, weight and Body Mass Index (BMI) is critical not only for effective clinical monitoring, but also for early detection of risk of cardiovascular disease (Cardiovascular Disease, CVD), personalized preventive strategies and customized treatment regimens. While these parameters are important, current methods for obtaining these parameters are limited. Manual anthropometric measurements can be subject to human error, operator-to-operator variation, and device-specific inconsistencies, which make them unreliable in a large-scale or telemedicine environment. Invasive or laboratory-based measurements, while accurate, are impractical for routine monitoring due to their associated discomfort, risk and time requirements. Even though Electrocardiograph (ECG) and photoplethysmogram (Photoplethysmogram, PPG) are used in practice, their application is mainly limited to Blood Pressure (BP) estimation, arrhythmia detection or pulse analysis, without a defined framework for demographic or anthropometric prediction. In addition, many existing artificial intelligence (ARTIFICIAL INTELLIGENCE, AI) systems in the healthcare field have prejudice and fairness issues that lead to predicted quality imbalances among different sexes, ages, or subgroups of people. Accordingly, there is a need for an improved system to address the shortcomings of conventional methods for determining demographic and anthropometric parameters. Disclosure of Invention The present disclosure aims to provide a system and method for non-invasive cardiovascular health assessment that will overcome or at least substantially ameliorate at least some of the disadvantages of the prior art, or at least provide an alternative. According to a first aspect of the present disclosure, a computer-implemented method for non-invasive cardiovascular health assessment is provided. The method comprises the following steps: At a processor, receiving a physiological signal of a subject, wherein the physiological signal comprises at least an ECG signal and a PPG signal; processing the ECG signal and the PPG signal by means of an AI model executed by the processor, and One or more demographic and anthropometric parameters are estimated by the AI model based on processing of the ECG and PPG signals. In one example, the processing step includes: Pre-processing the received ECG and PPG signals; feature extraction of the preprocessed ECG signal and PPG signal, and Selecting one or more optimal features, and Wherein the demographic and anthropometric parameters are estimated based on the one or more optimal characteristics. In an example, the processing step further comprises performing temporal processing on the pre-processed ECG and PPG signals. In one example, the temporal processing includes capturing a temporal correlation and a spatial relationship between the one or more optimal features. In an example, the demographic parameters include age, gender, and the one or more anthropometric features include weight, height, and BMI. In an example, the method further comprises: calculating a cardiovascular age of the subject based on the estimated demographic and anthropometric parameters by the AI model, and Outputting the cardiovascular age, and/or comparing the calculated cardiovascular age to a physiological age of the subject to identify a risk of vascular aging. In one example, the method further includes tracking the estimated BMI or weight change over time to detect an abnormal trend and presenting the estimated BMI or weight change on a user interface. In an example, the method further comprises: generating a visual indicator to show an estimate of the demographic and anthropometric parameters, and Presenting the visual indicator on a user interface, and Wherein the visual indicators include one or more of a saliency map, a heat map, or an attention heat map. In an example, the AI model includes a recurrent neural network (Recurrent Neural Network, RNN) model, wherein the RNN model is trained to perform the step of processing the received ECG and PPG signals and to output an estimate of one or more demographic and anthropometric parameters. In an example, the method may optionally apply an integration method to combine predictions of different model architectures. This may improve robustness and reliability. According to another aspect of the present disclosure, a system for non-invasive cardiovas