A traditional chinese medicine cardiovascular syndrome quantitative diagnosis system based on multi-modal physiological signals

Abstract

A traditional Chinese medicine cardiovascular syndrome quantitative diagnosis system based on multi-modal physiological signals, comprising: a multi-modal physiological signal synchronous acquisition and preprocessing module, integrating a radial artery pressure sensor array, an electrocardiogram electrode and a photoplethysmogram sensor, synchronously acquiring pulse waveforms, electrocardiograms and photoplethysmogram signals with a unified clock source; a multi-modal data deep fusion and feature extraction module, adopting a multi-branch attention fusion network, extracting deep features of pulse, electrocardiogram and photoplethysmogram through one-dimensional convolutional neural network, recurrent neural network and pulse wave transmission velocity calculation respectively, and adaptively weighting and fusing through cross-modal attention mechanism to obtain high-dimensional comprehensive features related to qi deficiency, blood stasis and phlegm-dampness; a traditional Chinese medicine syndrome quantitative scoring and diagnosis determination module. The objective quantitative evaluation of traditional Chinese medicine cardiovascular syndrome is realized, the discrimination accuracy of complex syndromes is significantly improved, and the portable and wearable application prospect is possessed.

Description

Technical Field

[0001] This invention relates to the fields of traditional Chinese medicine diagnostics, modern biomedical signal processing and artificial intelligence, and in particular to a quantitative diagnostic system for cardiovascular syndromes in traditional Chinese medicine based on multimodal physiological signals. Background Technology

[0002] Pulse diagnosis in Traditional Chinese Medicine (TCM) is the core of the four diagnostic methods of "inspection, auscultation, inquiry, and palpation," playing a crucial role in determining the balance of Qi, blood, Yin, and Yang, the strength of organ function, and the deficiency or excess of syndromes. Traditional pulse diagnosis relies on the physician's fingers to perceive the location, rate, shape, and force of the radial artery pulse. It is highly subjective and experience-dependent, leading to significant differences in diagnosis among different physicians. This makes it difficult to standardize, regulate, and promote its transmission, resulting in a lack of unified, objective, and quantitative standards for the diagnosis of cardiovascular syndromes in TCM.

[0003] With the development of sensor technology and artificial intelligence, researchers are attempting to quantify TCM syndromes by collecting multimodal physiological signals. For example, existing technologies acquire multimodal physiological signals by recording TCM consultations with patients, and perform retrospective analysis of the signals based on the consultation symptoms to determine the duration of symptom retrospective analysis. This allows for optimization of indicators based on their continuous impact, combined with Western medicine diagnostic data for comprehensive syndrome differentiation. However, dedicated multimodal deep feature extraction and quantitative scoring models have not been developed for core cardiovascular syndromes (Qi deficiency, blood stasis, phlegm dampness), limiting their ability to differentiate complex syndromes (such as Qi deficiency and blood stasis, or phlegm and blood stasis). Furthermore, the reliance on superficial feature splicing of tongue images and pulse signals fails to meet the clinical needs for dynamic assessment of syndrome severity and precise differentiation of complex syndromes. Existing multimodal data fusion methods largely remain at the level of feature splicing or decision voting, lacking a deep collaborative fusion architecture designed based on the holistic concept of Traditional Chinese Medicine (TCM). This fails to fully exploit the temporal synergy and nonlinear coupling characteristics among ECG, pulse, and PPG. Furthermore, for the quantitative assessment of core cardiovascular syndromes (Qi deficiency, blood stasis, and phlegm-dampness) and their complex syndromes, there is a lack of repeatable and standardized continuous scoring models, still relying on qualitative or binary classification outputs, which is insufficient to meet the quantitative monitoring needs in clinical efficacy evaluation and chronic disease management. In summary, current technologies cannot achieve high-precision, continuous quantitative, and repeatable diagnosis of TCM cardiovascular Qi deficiency, blood stasis, and phlegm-dampness syndromes, thus indicating significant room for improvement. Summary of the Invention

[0004] In order to overcome the above-mentioned shortcomings of the prior art, the purpose of this invention is to provide a quantitative diagnostic system for cardiovascular syndromes in traditional Chinese medicine based on multimodal physiological signals.

[0005] The technical solution adopted by this invention to solve its technical problem is: a quantitative diagnostic system for cardiovascular syndromes in Traditional Chinese Medicine based on multimodal physiological signals, comprising:

[0006] Multimodal physiological signal synchronous acquisition and preprocessing module: used to integrate a high-precision radial artery pressure sensor array, a standard ECG electrode module and a fingertip or earlobe photoplethysmography pulse wave sensor, and output time-aligned multimodal datasets;

[0007] Multimodal data deep fusion and feature extraction module: including pulse branch, ECG branch, PPG branch, and extracts high-dimensional comprehensive features related to Qi deficiency, blood stasis, and phlegm dampness through a multi-branch attention fusion network;

[0008] The TCM syndrome quantification scoring and diagnosis module is used to input the high-dimensional comprehensive features into a pre-trained machine learning regression model or classification model, output a continuous quantitative score of 0 to 10 for each of the three syndromes (Qi deficiency, blood stasis, and phlegm dampness), output the positive or negative syndrome judgment result according to the preset threshold, and output the judgment result of the compound syndrome according to the joint judgment rule.

[0009] As a further improvement of the present invention: the multimodal physiological signal synchronous acquisition and preprocessing module includes synchronous acquisition of radial artery pulse waveform, electrocardiogram signal and photoplethysmography pulse wave signal.

[0010] As a further improvement of the present invention, the multimodal physiological signal synchronous acquisition and preprocessing module includes time alignment, bandpass filtering, wavelet denoising, baseline correction and amplitude normalization or Z-score standardization processing of the acquired signals.

[0011] As a further improvement of the present invention: the multimodal physiological signal synchronous acquisition and preprocessing module includes:

[0012] Sensor array unit for acquiring radial artery pulse waveforms at a sampling frequency of ≥1000Hz;

[0013] ECG electrode unit, used for synchronous acquisition of standard electrocardiogram signals;

[0014] The photoplethysmography (PPG) unit is used to synchronously acquire PPG signals from the fingertip or earlobe.

[0015] The clock synchronization unit is used to unify the clock source and ensure that the time synchronization error of each signal is ≤1ms.

[0016] The preprocessing unit is used to sequentially perform time alignment, bandpass filtering, wavelet denoising, baseline correction, and amplitude normalization or Z-score standardization on the signal.

[0017] As a further improvement of the present invention: the multimodal data deep fusion and feature extraction module includes:

[0018] Pulse branch: used to extract the position, number, shape, momentum features and waveform geometric features of the pulse using a one-dimensional convolutional neural network;

[0019] ECG branch: RNN / LSTM extracts features from RR interval sequences;

[0020] PPG branch: Calculate pulse wave velocity PWV=L / Δt, and extract rising edge, falling edge, reflection wave index, and volume change characteristics.

[0021] As a further improvement of the present invention: the waveform geometric features extracted by the pulse branch include the main wave height, diphtheria wave height or position, pulse width, and the ratio of the area under the main wave to the area of ​​the whole wave.

[0022] As a further improvement of the present invention: the RR interval sequence features extracted from the electrocardiogram branches include the standard deviation of the RR interval, the root mean square of the difference between adjacent RR intervals, the ratio of low-frequency power to high-frequency power, and the sample entropy.

[0023] As a further improvement of the present invention: the preset thresholds include: a positive threshold of ≥6.5 points for Qi deficiency and a positive threshold of ≥7.0 points for blood stasis; the joint judgment rule includes: when the Qi deficiency score is ≥6.5 points and the blood stasis score is ≥7.0 points, it is judged as Qi deficiency and blood stasis syndrome.

[0024] As a further improvement of the present invention: the multimodal physiological signal synchronous acquisition and preprocessing module is configured to acquire continuous signals for at least 3 minutes.

[0025] As a further improvement of the present invention, the system is integrated into a portable or wearable data collection device.

[0026] Compared with the prior art, the beneficial effects of the present invention are:

[0027] By employing a high-precision pressure sensor array, standard ECG electrodes, and photoplethysmography (PPG) sensors, multimodal physiological signals such as radial artery pulse, ECG, and PPG are simultaneously acquired. A unified clock synchronization control ensures the objectivity and repeatability of the data from the signal acquisition source. Based on this, a multi-branch attention fusion network automatically extracts deep features highly correlated with Qi deficiency, blood stasis, and phlegm-dampness, achieving an objective, standardized, and repeatable quantitative assessment of TCM cardiovascular syndromes.

[0028] The constructed multi-branch attention fusion network has dedicated feature extraction branches for pulse wave, electrocardiogram (ECG), and photoplethysmography (PPG) pulse wave signals: the pulse wave branch extracts waveform geometric features such as position, number, shape, and potential using a one-dimensional convolutional neural network; the ECG branch extracts the time-domain, frequency-domain, and nonlinear features of the RR interval sequence using a recurrent neural network or long short-term memory network; and the PPG pulse wave branch calculates the pulse wave conduction velocity and volume change characteristics. Based on this, an adaptive weighted fusion of features from each branch is performed through a cross-modal attention mechanism, which can fully explore the inherent synergistic and nonlinear coupling relationships among the three signals.

[0029] Independent continuous quantitative scoring models ranging from 0 to 10 points were established for the three core syndromes of Qi deficiency, blood stasis, and phlegm-dampness. Clear positive thresholds and joint judgment rules for complex syndromes were set, which can reflect the severity of the syndromes and provide quantifiable and comparable objective evidence for the dynamic evaluation of TCM clinical efficacy, scientific research, and chronic disease management.

[0030] Attached image description (It seems there are no accompanying labels; should we add them?)

[0031] To more clearly illustrate the technical solution, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is a schematic diagram of the structural framework of the present invention.

[0033] Figure 2 This is a schematic diagram of the hardware architecture for synchronous acquisition of multimodal physiological signals according to the present invention.

[0034] Figure 3 This is a schematic diagram of the multi-branch attention fusion network model structure of the present invention.

[0035] Figure 4 This is a schematic diagram of the quantitative scoring and diagnostic process for cardiovascular syndromes in Traditional Chinese Medicine according to the present invention. Detailed Implementation

[0036] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0037] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0038] Accurate identification of cardiovascular syndromes (such as Qi deficiency, blood stasis, and phlegm-dampness) in Traditional Chinese Medicine (TCM) is crucial for clinical intervention in diseases like coronary heart disease, heart failure, and hypertension. Traditional TCM relies on pulse diagnosis, depending on the physician's finger perception of the radial artery pulse's location, rate, shape, and momentum. However, this method is highly subjective, experience-dependent, and inconsistent among physicians, making standardization and objective quantification difficult. With the development of sensor and artificial intelligence technologies, current research attempts to assist TCM syndrome identification through multimodal physiological signals, but the following technical challenges remain: existing solutions mostly employ shallow fusion methods such as feature splicing or decision voting, lacking a deep, collaborative fusion architecture based on the holistic concept of TCM, failing to fully explore the inherent temporal synergy and nonlinear coupling characteristics between electrocardiogram (ECG), photoplethysmography (PPG), and radial artery pulse; existing systems mainly output qualitative or binary classification results, failing to establish a repeatable, standardized, and continuous quantitative scoring system for Qi deficiency, blood stasis, phlegm-dampness, and their combined syndromes (Qi deficiency and blood stasis, phlegm and blood stasis), making it difficult to meet the quantitative monitoring needs for dynamic clinical efficacy evaluation and chronic disease management. The existing system does not achieve millisecond-level synchronization during multimodal signal acquisition, which affects the accuracy of subsequent feature fusion and the reliability of quantitative diagnosis.

[0039] To address the aforementioned problems, this invention provides a quantitative diagnostic system for cardiovascular syndromes in Traditional Chinese Medicine based on multimodal physiological signals. The invention is further described below with reference to the accompanying drawings and embodiments: Figures 1-4 As shown, it includes:

[0040] Multimodal physiological signal synchronous acquisition and preprocessing module: used to integrate a high-precision radial artery pressure sensor array, a standard ECG electrode module and a fingertip or earlobe photoplethysmography pulse wave sensor, and output time-aligned multimodal datasets;

[0041] Multimodal data deep fusion and feature extraction module: including pulse branch, ECG branch, PPG branch, and extracts high-dimensional comprehensive features related to Qi deficiency, blood stasis, and phlegm dampness through a multi-branch attention fusion network;

[0042] The TCM syndrome quantification scoring and diagnosis module is used to input the high-dimensional comprehensive features into a pre-trained machine learning regression model or classification model, output a continuous quantitative score of 0 to 10 for each of the three syndromes (Qi deficiency, blood stasis, and phlegm dampness), output the positive or negative syndrome judgment result according to the preset threshold, and output the judgment result of the compound syndrome according to the joint judgment rule.

[0043] As one embodiment of the present invention, a quantitative diagnostic system for cardiovascular syndromes in Traditional Chinese Medicine based on multimodal physiological signals includes:

[0044] 1) Multimodal physiological signal synchronous acquisition and preprocessing module; 2) Multimodal data deep fusion and feature extraction module; 3) Traditional Chinese medicine syndrome quantitative scoring and diagnostic judgment module.

[0045] Multimodal physiological signal synchronous acquisition and preprocessing module: integrates a high-precision radial artery pressure sensor array, a standard ECG electrode module, and a fingertip / earlobe photoelectric PPG sensor, with a unified clock source for synchronous triggering and a time synchronization error of ≤1ms;

[0046] Simultaneous acquisition: radial artery pulse waveform, ECG electrocardiogram signal, PPG photoplethysmography (PPG) pulse wave signal;

[0047] Preprocessing includes: temporal alignment, bandpass filtering, wavelet denoising, baseline correction, amplitude normalization / Z-score standardization, outputting a high-quality temporally aligned multimodal dataset.

[0048] Multimodal data deep fusion and feature extraction module: Employs a multi-branch attention fusion network.

[0049] Pulse branching: One-dimensional CNN extracts position, number, shape, momentum and waveform geometric features (main wave height, diphtheria wave, width, area ratio, etc.);

[0050] ECG branch: RNN / LSTM extracts RR interval sequence features (SDNN, RMSSD, LF / HF, sample entropy, etc.).

[0051] PPG branch: Calculate pulse wave velocity PWV=L / Δt, and extract rising edge, falling edge, reflection wave index, and volume change characteristics;

[0052] By using cross-modal attention mechanism for adaptive weighted fusion, high-dimensional comprehensive features related to Qi deficiency, blood stasis, and phlegm-dampness are extracted.

[0053] The TCM syndrome quantitative scoring and diagnostic judgment module: Based on a large number of TCM physician-annotated clinical samples, a machine learning regression / classification model is trained to establish independent quantitative scoring models and thresholds for Qi deficiency, blood stasis, and phlegm dampness, respectively. The module constructs joint judgment rules for complex syndromes (Qi deficiency and blood stasis, phlegm and blood stasis, etc.) and outputs a continuous quantitative score of 0-10 points and a positive / negative judgment result for the syndrome.

[0054] As a specific embodiment of the present invention, multimodal signal synchronous acquisition: a portable acquisition device is built, integrating a pulse pressure sensor (≥1000Hz), a standard ECG electrode, and a fingertip PPG sensor, with unified clock synchronization and an error ≤1ms; the subject sits still, and three signals are acquired for 3 minutes.

[0055] Signal preprocessing: The following steps are performed in sequence: timestamp alignment → wavelet denoising → removal of power frequency interference and motion artifacts → baseline correction → amplitude normalization, to obtain a standard multimodal dataset.

[0056] Multi-branch attention feature fusion: Pulse image: One-dimensional CNN extracts position, number, shape, and potential features;

[0057] ECG: Calculate RR intervals and extract SDNN, RMSSD, and sample entropy;

[0058] PPG: Calculate PWV=L / Δt and extract rise time and reflection index;

[0059] The comprehensive feature vector is obtained by weighted fusion through a cross-modal attention layer.

[0060] Syndrome Quantification Scoring and Diagnosis: A gradient boosting tree / support vector regression model is trained using labeled clinical samples to output quantitative values ​​of 0–10 for Qi deficiency, blood stasis, and phlegm-dampness, respectively. Thresholds are set: for example, Qi deficiency ≥ 6.5 points and blood stasis ≥ 7.0 points are considered positive. Complex syndromes are determined according to combined rules (e.g., Qi deficiency ≥ 6.5 and blood stasis ≥ 7.0 constitute Qi deficiency and blood stasis syndrome). The system automatically outputs quantitative scores and syndrome differentiation results, achieving objective and standardized diagnosis of cardiovascular syndromes in Traditional Chinese Medicine.

[0061] Working principle of the invention:

[0062] Through the coordinated operation of three core modules, the system achieves simultaneous acquisition of multimodal physiological signals, deep feature fusion, and quantitative assessment of cardiovascular syndromes in Traditional Chinese Medicine.

[0063] (1) High-precision synchronous acquisition and preprocessing of multimodal physiological signals:

[0064] The system integrates a high-precision radial artery pressure sensor array, a standard ECG electrode module, and a fingertip or earlobe photoplethysmography (PPG) sensor. All three sensors are synchronously triggered by a unified clock source, with a time synchronization error controlled within 1ms, and synchronously acquire the following three signals:

[0065] Radial artery pulse waveform: reflects the original mechanical information of the pulse, such as its location, rate, shape, and momentum;

[0066] Electrocardiogram (ECG) signals: reflect the temporal changes in cardiac electrophysiological activity;

[0067] Photoplethysmography (PPG) signal: reflects changes in peripheral vascular volume pulse wave.

[0068] The acquired raw signals are sequentially processed through time alignment, bandpass filtering, wavelet denoising, baseline correction, and amplitude normalization (or Z-score standardization) to eliminate noise, motion artifacts, and individual differences, outputting a high-quality, time-aligned multimodal dataset.

[0069] (2) Deep feature extraction of multi-branch attention fusion network:

[0070] The preprocessed multimodal data is input into a multi-branch attention fusion network, which contains three parallel branches and a cross-modal attention fusion layer:

[0071] Pulse branching: Using a one-dimensional convolutional neural network (1D-CNN), the position (pulse location), number (pulse frequency), shape (waveform morphology), momentum (pulse strength), and waveform geometric features of the pulse signal are automatically extracted, including the height of the main wave, the height or position of the dicrotic wave, the width of the pulse map, and the ratio of the area under the main wave to the area of ​​the entire wave.

[0072] Electrocardiogram (ECG) branch: Recurrent neural network (RNN) or long short-term memory network (LSTM) is used to extract RR interval sequence features, including RR interval standard deviation (SDNN), root mean square of the difference between adjacent RR intervals (RMSSD), low-frequency power to high-frequency power ratio (LF / HF), sample entropy and other time-domain, frequency-domain and nonlinear features.

[0073] Photoplethysmography (PPG) branching: Calculate the pulse wave propagation velocity (PWV = L / Δt, where L is the propagation distance from the ECG R wave peak to the rising edge of the PPG, and Δt is the propagation time difference), and extract waveform features such as rising edge time, falling edge time, reflection wave index, and volume change.

[0074] Cross-modal attention mechanism: For the feature vectors output by the above three branches, the attention weights are calculated in the channel or spatial dimension, and the weighted fusion is adaptively performed to highlight the synergistic features that are highly correlated with Qi deficiency, blood stasis and phlegm dampness, suppress irrelevant or redundant information, and finally output a high-dimensional comprehensive feature vector.

[0075] (3) Quantitative scoring and judgment of cardiovascular syndromes in Traditional Chinese Medicine:

[0076] The high-dimensional comprehensive feature vector is input into a pre-trained machine learning model (gradient boosting tree or support vector regression model). This model is trained based on a large number of clinical samples labeled by TCM doctors and can map the severity of three independent syndromes: Qi deficiency, blood stasis, and phlegm dampness, and output a continuous quantitative score of 0 to 10 (0 points indicates no symptoms of the syndrome, and 10 points indicates extremely severe symptoms).

[0077] The system automatically determines the positive result based on a preset positive threshold:

[0078] Positive result for Qi deficiency syndrome: Qi deficiency score ≥ 6.5 points;

[0079] Positive blood stasis syndrome: Blood stasis score ≥ 7.0;

[0080] Positive phlegm-dampness syndrome: Phlegm-dampness score (based on the corresponding threshold set according to the training samples).

[0081] At the same time, the system outputs the results of the complex syndrome determination according to the rules for joint determination of complex syndromes, for example:

[0082] If the Qi deficiency score is ≥ 6.5 and the blood stasis score is ≥ 7.0, then it is determined to be "Qi deficiency and blood stasis syndrome";

[0083] If both the phlegm-dampness score and the blood stasis score reach the corresponding threshold, it is determined to be "phlegm-blood stasis syndrome".

[0084] Ultimately, the system outputs quantitative scores, positive / negative judgment results, and complex syndrome types for Qi deficiency, blood stasis, and phlegm dampness in a visual manner, providing a quantitative basis for the objective and standardized assessment of cardiovascular syndromes in traditional Chinese medicine.

[0085] The main functions of this invention are:

[0086] This system integrates a radial artery pressure sensor array, standard ECG electrodes, and fingertip / earlobe photoplethysmography (PPG) sensors to synchronously trigger the acquisition of pulse waveforms, ECG, and PPG signals using a unified clock source. The time synchronization error is ≤1ms, and the system features signal preprocessing capabilities, outputting high-quality multimodal datasets. Based on a multi-branch attention fusion network, specific features are extracted from the pulse, ECG, and PPG signals respectively. Adaptive weighted fusion via a cross-modal attention mechanism generates high-dimensional comprehensive feature vectors related to Qi deficiency, blood stasis, and phlegm-dampness. Using a pre-trained machine learning regression / classification model, continuous quantitative scores of 0-10 are output for each of the three independent syndromes (Qi deficiency, blood stasis, and phlegm-dampness) based on the fused features, enabling precise quantitative assessment of syndrome severity. The system automatically determines the positive or negative results for each syndrome based on preset thresholds and outputs the judgment results for composite syndromes according to joint judgment rules. The system can be integrated into portable or wearable acquisition devices, suitable for various scenarios such as clinics, homes, and remote monitoring, meeting the clinical and daily health management needs for objective assessment of TCM cardiovascular syndromes.

[0087] In summary, after reading this invention document, those skilled in the art can make various other corresponding modifications to the technical solutions and concepts based on this invention without creative mental effort, and all of these modifications fall within the scope of protection of this invention.

Claims

1. A quantitative diagnostic system for cardiovascular syndromes in Traditional Chinese Medicine based on multimodal physiological signals, characterized in that, include: Multimodal physiological signal synchronous acquisition and preprocessing module: used to integrate a high-precision radial artery pressure sensor array, a standard ECG electrode module and a fingertip or earlobe photoplethysmography pulse wave sensor, and output time-aligned multimodal datasets; Multimodal data deep fusion and feature extraction module: including pulse branch, ECG branch, PPG branch, and extracts high-dimensional comprehensive features related to Qi deficiency, blood stasis, and phlegm dampness through a multi-branch attention fusion network; The TCM syndrome quantification scoring and diagnosis module is used to input the high-dimensional comprehensive features into a pre-trained machine learning regression model or classification model, output a continuous quantitative score of 0 to 10 for each of the three syndromes (Qi deficiency, blood stasis, and phlegm dampness), output the positive or negative syndrome judgment result according to the preset threshold, and output the judgment result of the compound syndrome according to the joint judgment rule.

2. The quantitative diagnostic system for cardiovascular syndromes in Traditional Chinese Medicine based on multimodal physiological signals according to claim 1, characterized in that, The multimodal physiological signal synchronous acquisition and preprocessing module includes synchronous acquisition of radial artery pulse waveform, electrocardiogram signal and photoplethysmography pulse wave signal.

3. The quantitative diagnostic system for cardiovascular syndromes in Traditional Chinese Medicine based on multimodal physiological signals according to claim 2, characterized in that, The multimodal physiological signal synchronous acquisition and preprocessing module includes time alignment, bandpass filtering, wavelet denoising, baseline correction, and amplitude normalization or Z-score standardization of the acquired signals.

4. The quantitative diagnostic system for cardiovascular syndromes in Traditional Chinese Medicine based on multimodal physiological signals according to claim 3, characterized in that, The multimodal physiological signal synchronous acquisition and preprocessing module includes: Sensor array unit for acquiring radial artery pulse waveforms at a sampling frequency of ≥1000Hz; ECG electrode unit, used for synchronous acquisition of standard electrocardiogram signals; The photoplethysmography (PPG) unit is used to synchronously acquire PPG signals from the fingertip or earlobe. The clock synchronization unit is used to unify the clock source and ensure that the time synchronization error of each signal is ≤1ms. The preprocessing unit is used to sequentially perform time alignment, bandpass filtering, wavelet denoising, baseline correction, and amplitude normalization or Z-score standardization on the signal.

5. The quantitative diagnostic system for cardiovascular syndromes in Traditional Chinese Medicine based on multimodal physiological signals according to claim 1, characterized in that, The multimodal data deep fusion and feature extraction module includes: Pulse branch: used to extract the position, number, shape, momentum features and waveform geometric features of the pulse using a one-dimensional convolutional neural network; ECG branch: RNN / LSTM extracts features from RR interval sequences; PPG branch: Calculate pulse wave velocity PWV=L / Δt, and extract rising edge, falling edge, reflection wave index, and volume change characteristics.

6. The quantitative diagnostic system for cardiovascular syndromes in Traditional Chinese Medicine based on multimodal physiological signals according to claim 5, characterized in that, The waveform geometric features extracted from the pulse branch include the main wave height, diphtheria wave height or position, pulse width, and the ratio of the area under the main wave to the area of ​​the entire wave.

7. The quantitative diagnostic system for cardiovascular syndromes in Traditional Chinese Medicine based on multimodal physiological signals according to claim 1, characterized in that, The RR interval sequence features extracted from the electrocardiogram branches include the standard deviation of the RR interval, the root mean square of the difference between adjacent RR intervals, the ratio of low-frequency power to high-frequency power, and the sample entropy.

8. The quantitative diagnostic system for cardiovascular syndromes in Traditional Chinese Medicine based on multimodal physiological signals according to claim 1, characterized in that, The preset thresholds include: a positive threshold of ≥6.5 for Qi deficiency, ≥7.0 for blood stasis, and ≥6.0 for phlegm-dampness. The combined judgment rules include: when the Qi deficiency score is ≥6.5 and the blood stasis score is ≥7.0, it is judged as Qi deficiency and blood stasis syndrome. When the blood stasis score is ≥7.0 and the phlegm-dampness score is ≥6.0, it is judged as phlegm-stasis syndrome.

9. A quantitative diagnostic system for cardiovascular syndromes in Traditional Chinese Medicine based on multimodal physiological signals according to claim 1, characterized in that, The multimodal physiological signal synchronous acquisition and preprocessing module is configured to acquire continuous signals for at least 3 minutes.

10. A quantitative diagnostic system for cardiovascular syndromes in Traditional Chinese Medicine based on multimodal physiological signals according to claim 1, characterized in that, The system is integrated into a portable or wearable data collection device.

Images