Traditional Chinese medicine physique intelligent typing method and system for carotid plaque patient

Abstract

The invention discloses a traditional Chinese medicine physique intelligent typing method and system for a carotid plaque patient, and relates to the technical field of artificial intelligence, the method comprises the following steps: correlating heart rate variability characteristic parameters, basic information characteristics and traditional Chinese medicine physique type identification results, and constructing a training data set; constructing a decision tree model; the method comprises the following steps: acquiring to-be-detected face video data and to-be-detected basic information data of a to-be-detected carotid plaque patient, and extracting to-be-detected heart rate variability characteristic parameters and to-be-detected basic information characteristics corresponding to the to-be-detected face video data and the to-be-detected basic information data; and inputting the to-be-detected heart rate variability characteristic parameters and the to-be-detected basic information characteristics into the decision tree model to obtain a traditional Chinese medicine constitution type identification result corresponding to the to-be-detected carotid plaque patient, thereby facilitating solving the problem that the traditional Chinese medicine constitution intelligent typing of the carotid plaque patient cannot be realized in the prior art.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method and system for intelligent TCM constitution classification of patients with carotid artery plaque. Background Art

[0002] The carotid arteries are crucial for blood supply to the brain. Plaque buildup within these arteries can lead to vascular stenosis. When stenosis exceeds halfway, blood flow to the brain is reduced, causing symptoms such as dizziness. Dislodged blood clots can also cause life-threatening cerebral infarction. Preventing and treating carotid atherosclerotic plaques has become a medical challenge.

[0003] Carotid ultrasound is noninvasive and convenient, making it the preferred diagnostic method, revealing the presence of coronary artery disease. Western medicine primarily treats atherosclerotic conditions like carotid plaques with lifestyle interventions and lipid-lowering measures, lacking more targeted approaches. Traditional Chinese Medicine emphasizes a holistic approach, but its understanding of carotid plaques and other conditions is imprecise. Following the release of the "Classification and Determination of Chinese Constitutions in Traditional Chinese Medicine" in 2009, research on the distribution of diseases by TCM constitution has flourished. While TCM constitution theory can provide new insights into disease diagnosis and treatment, existing technologies lack a method for intelligently categorizing the TCM constitution of patients with carotid plaques.

[0004] Therefore, there is an urgent need for a TCM constitution intelligent classification method for patients with carotid artery plaques. Summary of the Invention

[0005] In view of this, the present invention proposes a method and system for intelligent TCM constitution typing of patients with carotid artery plaques, which can realize intelligent TCM constitution typing of patients with carotid artery plaques.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A method for intelligent TCM constitution classification of patients with carotid artery plaque, comprising:

[0008] Collect facial video data and basic information data of patients with carotid artery plaque;

[0009] Extracting basic information features corresponding to the basic information data;

[0010] extracting a photoplethysmography signal corresponding to the facial video data;

[0011] Extracting heart rate variability characteristic parameters corresponding to the photoplethysmography signal;

[0012] In response to a request to obtain a constitution type of a patient with carotid artery plaque, conducting a questionnaire survey on the patient with carotid artery plaque to obtain a result of identifying a TCM constitution type corresponding to the patient with carotid artery plaque;

[0013] Correlate the heart rate variability characteristic parameters, basic information characteristics and TCM constitution type identification results to construct a training data set;

[0014] Constructing a decision tree model, and training the decision tree model based on the training data set to obtain an optimal decision tree model;

[0015] Acquire facial video data of the face to be tested and basic information data to be tested of the patient with carotid artery plaque to be tested, extract characteristic parameters of heart rate variability to be tested and basic information features to be tested corresponding to the facial video data to be tested and the basic information data to be tested, and input the characteristic parameters of heart rate variability to be tested and the basic information features to be tested into the optimal decision tree model to obtain a result of identifying the TCM constitution type corresponding to the patient with carotid artery plaque to be tested.

[0016] On the basis of the above technical solution, the present invention can also be improved as follows:

[0017] Optionally, extracting a photoplethysmography signal corresponding to the facial video data includes:

[0018] Obtaining facial single-channel pixels corresponding to the facial video data of a human face, and obtaining a light change curve corresponding to the facial video data of the human face based on the facial single-channel pixels;

[0019] A fast independent component analysis algorithm is used to obtain a source signal with the highest Pearson coefficient with the green channel, and the source signal is used as a photoplethysmography signal related to the heart beat;

[0020] The photoplethysmography signal is processed by wavelet transform, fast Fourier transform and narrowband filtering to obtain a pure photoplethysmography signal.

[0021] Optionally, extracting a heart rate variability characteristic parameter corresponding to the photoplethysmography signal includes:

[0022] After the pure photoplethysmography signal is up-sampled by cubic spline interpolation, the peak points are extracted and the time intervals between the peak points are calculated to obtain the RR interval series.

[0023] Acquire an HRV time domain curve based on the RR interval sequence, and extract a time domain characteristic parameter corresponding to the photoplethysmography signal based on the HRV time domain curve;

[0024] The Welch power spectrum is used to extract the frequency domain characteristic parameters corresponding to the photoplethysmography signal;

[0025] The nonlinear characteristic parameters corresponding to the photoplethysmography signal are extracted based on the Poincare scatter plot analysis method;

[0026] A heart rate variability characteristic parameter is obtained based on the time domain characteristic parameter, the frequency domain characteristic parameter and the nonlinear characteristic parameter.

[0027] Optionally, the TCM constitution type identification results include balanced constitution, qi deficiency constitution, yang deficiency constitution, yin deficiency constitution, blood stasis constitution, phlegm-damp constitution, damp-heat constitution, qi stagnation constitution and special constitution.

[0028] Optionally, the constructing of the decision tree model, training the decision tree model based on the training data set to obtain an optimal decision tree model, includes:

[0029] Calculate the loss function value by formula (1);

[0030] Gain(A)=Info(D)-Info A (D) Formula (1);

[0031] Where Gain(A) is the information gain value, Info(D) is the information entropy of the data set D, and Info A (D) is the information entropy obtained after dividing the data set D according to attribute A;

[0032] The attribute feature with the largest information gain value Gain(A) is used as the root node, and the child nodes are split from the root node. The information gain is calculated and the child node with the largest information gain is used as the child node until all attribute features are less than the set threshold or no attribute feature is selected.

[0033] Optionally, the calculation of the loss function value by formula (1) further includes:

[0034] Calculate the information entropy of data set D using formula (2);

[0035]

[0036] Where Info(D) is the information entropy of the dataset D, c is the number of nine TCM constitution categories corresponding to the carotid artery plaque patient samples, and p i is the proportion of samples belonging to category i to all samples.

[0037] Optionally, the calculation of the loss function value by formula (1) further includes:

[0038] The information entropy obtained by dividing the data set D according to attribute A is calculated by formula (3);

[0039]

[0040] In the formula, Info A (D) is the information entropy obtained after dividing the data set D according to attribute A, k is the number of possible values ​​of attribute A, Dj is the subset corresponding to the jth value of attribute A, |D j | is a subset D j The number of samples in dataset D, |D| is the total number of samples in dataset D.

[0041] An intelligent TCM constitution classification system for patients with carotid artery plaque, comprising:

[0042] An acquisition module, used to collect facial video data and basic information data of patients with carotid artery plaque;

[0043] A first extraction module is used to extract basic information features corresponding to the basic information data;

[0044] a second extraction module, configured to extract a photoplethysmography signal corresponding to the facial video data;

[0045] a third extraction module, configured to extract characteristic parameters of heart rate variability corresponding to the photoplethysmography signal;

[0046] a survey module, configured to respond to a request for obtaining a constitution type of a carotid artery plaque patient and conduct a questionnaire survey on the carotid artery plaque patient to obtain a result of identifying a TCM constitution type corresponding to the carotid artery plaque patient;

[0047] A training set construction module is used to associate heart rate variability characteristic parameters, basic information characteristics and TCM constitution type identification results to construct a training data set;

[0048] A model building module is used to build a decision tree model and train the decision tree model based on the training data set to obtain an optimal decision tree model;

[0049] The result acquisition module is used to obtain facial video data of the face to be tested and basic information data to be tested of the patient with carotid artery plaque to be tested, extract the heart rate variability characteristic parameters to be tested and the basic information features to be tested corresponding to the facial video data to be tested and the basic information data to be tested, and input the heart rate variability characteristic parameters to be tested and the basic information features to be tested into the optimal decision tree model to obtain the TCM constitution type identification result corresponding to the patient with carotid artery plaque to be tested.

[0050] An electronic device comprises a memory, a processor and a computer program stored in the memory and running on the processor, wherein the steps of the method are implemented when the processor executes the computer program.

[0051] A non-transitory computer-readable storage medium stores a computer program, which implements the steps of the method when executed by a processor.

[0052] The present invention has the following advantages:

[0053] The intelligent TCM constitution classification method for patients with carotid artery plaques in the present invention integrates heart rate variability characteristic parameters, basic information characteristics and questionnaire survey results extracted from facial video data, integrates multi-dimensional data, gets rid of simple subjective judgment, makes the TCM constitution classification of patients with carotid artery plaques more objective and accurate, and reduces human errors.

[0054] The intelligent TCM constitution classification method for patients with carotid artery plaques in the present invention constructs and trains a decision tree model, which can quickly process large amounts of data and realize automated TCM constitution classification. Compared with traditional manual classification, the efficiency is greatly improved and can adapt to large-scale clinical applications and research needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] For purposes of illustration and not limitation, the present invention will now be described with reference to embodiments thereof and the accompanying drawings, in which:

[0056] Figure 1 Schematic diagram of the process of the TCM constitution intelligent typing method for patients with carotid artery plaque in an embodiment of the present invention;

[0057] Figure 2 Schematic diagram of the main components of the TCM constitution intelligent classification system for patients with carotid artery plaques according to an embodiment of the present invention;

[0058] Figure 3 This is a schematic diagram of the physical structure of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0059] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of the present invention.

[0060] It should be noted that the terms "first," "second," and the like in the description of the present invention and the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a particular order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate for the embodiments of the present invention described herein. In addition, the terms "including," "having," and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to these processes, methods, products, or apparatuses.

[0061] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features thereof can be combined with each other. The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0062] Figure 1 FIG. 1 is a flow chart of a method for intelligent TCM constitution typing of patients with carotid artery plaque according to an embodiment of the present invention. Figure 1 As shown, the TCM constitution intelligent typing method for patients with carotid artery plaque provided by an embodiment of the present invention includes the following steps S101 to S108.

[0063] S101, collecting facial video data and basic information data of patients with carotid artery plaque.

[0064] A camera is used to capture facial video data from the subject. The human face is rich in capillaries and is located near the carotid artery, resulting in high blood signal intensity. Video capture requirements include uniform and sufficiently bright lighting in the acquisition environment (to ensure good light quality reflected from the skin surface), a high camera resolution (to ensure a clear facial image), appropriate camera position and angle (the camera is positioned at an appropriate distance from the subject's face and perpendicular to the face), the subject maintaining a relatively still posture (to ensure stable video), and a video length of at least 15 seconds (it is recommended to capture longer videos to ensure sufficient data for signal processing and analysis). Furthermore, external interference such as background noise, camera shake, and occlusions should be minimized. In summary, effective RPPG video capture requires careful consideration of lighting conditions, camera selection and placement, subject posture, and other factors, minimizing interference to obtain high-quality physiological signal data.

[0065] S102: Extract basic information features corresponding to the basic information data.

[0066] Basic information data includes patient age, gender, body mass index, waist-to-hip ratio, blood sugar level, and blood lipid level.

[0067] S103, extracting the photoplethysmography signal corresponding to the facial video data.

[0068] Obtaining facial single-channel pixels corresponding to the facial video data of a human face, and obtaining a light change curve corresponding to the facial video data of the human face based on the facial single-channel pixels;

[0069] A fast independent component analysis algorithm is used to obtain a source signal with the highest Pearson coefficient with the green channel, and the source signal is used as a photoplethysmography signal related to the heart beat;

[0070] The photoplethysmography signal is processed by wavelet transform, fast Fourier transform and narrowband filtering to obtain a pure photoplethysmography signal.

[0071] S104: extracting heart rate variability characteristic parameters corresponding to the photoplethysmography signal.

[0072] After the pure photoplethysmography signal is up-sampled by cubic spline interpolation, the peak points are extracted and the time intervals between the peak points are calculated to obtain the RR interval series.

[0073] Acquire an HRV time domain curve based on the RR interval sequence, and extract a time domain characteristic parameter corresponding to the photoplethysmography signal based on the HRV time domain curve;

[0074] The Welch power spectrum is used to extract the frequency domain characteristic parameters corresponding to the photoplethysmography signal;

[0075] The nonlinear characteristic parameters corresponding to the photoplethysmography signal are extracted based on the Poincare scatter plot analysis method;

[0076] A heart rate variability characteristic parameter is obtained based on the time domain characteristic parameter, the frequency domain characteristic parameter and the nonlinear characteristic parameter.

[0077] S105 , in response to a request to obtain the constitution type of a carotid artery plaque patient, conducting a questionnaire survey on the carotid artery plaque patient to obtain a TCM constitution type identification result corresponding to the carotid artery plaque patient.

[0078] A questionnaire and survey of the aforementioned carotid artery plaque patients, using Professor Wang Qi's "Nine-Point Constitution Scale," yielded results from a Traditional Chinese Medicine (TCM) constitution identification process: nine types of constitution: balanced constitution, Qi deficiency, Yang deficiency, Yin deficiency, blood stasis, phlegm-dampness, damp-heat, Qi stagnation, and special constitution. For example, a questionnaire survey of 206 patients with carotid artery plaque revealed that the constitutions of these patients varied between simple and complex types, with the proportions of these two types roughly equal. Among these, 101 patients had a simple constitution and 105 had a complex constitution. The eight most common constitution types were phlegm-dampness, Qi deficiency and phlegm-dampness, Qi deficiency and blood stasis, Qi deficiency, blood stasis, balanced constitution, Qi deficiency and yang deficiency, and Yang deficiency and phlegm-dampness, accounting for 68.4% of the total population.

[0079] S106 , correlating the heart rate variability characteristic parameters, basic information characteristics, and TCM constitution type identification results to construct a training data set.

[0080] Heart rate variability characteristic parameters include:

[0081] SDNN: standard deviation of heart rate interval; RMSSD: root mean square of the sum of squares of the differences between adjacent normal heart rate intervals; LF: amplitude of normal heart rate intervals in the low frequency range (0.04-0.15 Hz); HF: amplitude of normal heart rate intervals in the high frequency range (0.15-40 Hz); LF / HF: low-high frequency power ratio; sdsd: standard deviation of adjacent RR differences; hf_nu: normalized high frequency power; lf_nu: normalized low frequency power; nn: normal heart rate interval, RR interval; cvnni: coefficient of variation, standard deviation of the series data of heart beat variability; nn20: number of adjacent RR interval differences > 20 ms; nn50: number of adjacent RR interval differences > 50 ms; pnn20: NN20 (number of adjacent heart beat interval differences greater than 20 ms) =The following table lists the percentage of NN50 (number of adjacent heartbeat intervals with a difference greater than 50 milliseconds) in the total heartbeat number: pnn50: percentage of NN50 (number of adjacent heartbeat intervals with a difference greater than 50 milliseconds) in the total heartbeat number: cvcd: coefficient of variation of continuous difference: sd1: standard deviation of the projection of the Poincare plot on the line perpendicular to the equation: sd2: standard deviation of the Poincare projection: SDNNindex: mean of the standard deviations of all RR intervals: Hrviendex: total number of RR intervals divided by the height of the RR interval: Max_min: difference between the maximum and minimum values ​​of all RR intervals: MEAN: mean of the RR intervals: totalpower: total power spectral density: VLF: very low frequency power, ranging from 0.04 to 0.15 Hz: HFn: normalized high frequency capability: LFn: proportion of low frequency power in the total power spectrum:

[0082] Compared with the balanced constitution, the SDANN of the total biased constitution was lower (P < 0.05), indicating that the sympathetic nerve activity of the biased constitution is lower than that of the balanced constitution. The trends of changes in heart rate variability indicators varied among different biased constitutions: compared with the balanced constitution, the SDANN, VLF, and LF / HF of the yang deficiency constitution were lower (P < 0.05), and the HF was higher (P < 0.05). This indicates that the lower sympathetic-parasympathetic balance ratio LF / HF in the yang deficiency constitution is due to the combined effects of lower sympathetic nerve activity and higher parasympathetic nerve activity. The SDANN and SDNN of the qi deficiency and blood stasis constitutions were lower than those of the balanced constitution (P < 0.05), indicating that the sympathetic nerve activity is lower in the qi deficiency and blood stasis constitutions, and the overall regulation of heart rate by the autonomic nervous system is weakened. The VLF of the phlegm-damp constitution was lower (P < 0.05).

[0083] People with qi deficiency, phlegm-dampness, or blood stasis constitutions are more susceptible to carotid artery plaques. Male patients with carotid artery plaques are predominantly characterized by phlegm-dampness and qi deficiency-phlegm-dampness constitutions, while female patients are predominantly characterized by qi deficiency-blood stasis and blood stasis constitutions. Middle-aged patients are predominantly characterized by phlegm-dampness and qi deficiency-phlegm-dampness constitutions, while elderly patients are predominantly characterized by qi deficiency-blood stasis, qi deficiency, and blood stasis constitutions. Patients with phlegm-dampness constitutions have the highest BMI and WHR. Blood lipid levels in patients with carotid artery plaques are highly correlated with phlegm-dampness constitutions. There is no significant correlation between blood sugar and constitution in patients with carotid artery plaques.

[0084] S107, building a decision tree model, and training the decision tree model based on the training data set to obtain an optimal decision tree model.

[0085] Based on a given data set, the attribute feature with the maximum information gain is selected as the root node of a decision tree. Each child node in the decision tree represents a certain attribute feature data of a sample, and the leaf nodes of the decision tree represent the nine major TCM constitution categories to which the carotid artery plaque patient samples belong.

[0086] Calculate the loss function value by formula (1);

[0087] Gain(A)=Info(D)-Info A (D) Formula (1);

[0088] Where Gain(A) is the information gain value, Info(D) is the information entropy of the data set D, and Info A (D) is the information entropy obtained after dividing the data set D according to attribute A;

[0089] The attribute feature with the largest information gain value Gain(A) is used as the root node, and the child nodes are split from the root node. The information gain is calculated and the child node with the largest information gain is used as the child node until all attribute features are less than the set threshold or no attribute feature is selected.

[0090] Calculate the information entropy of data set D using formula (2);

[0091]

[0092] Where Info(D) is the information entropy of the dataset D, c is the number of nine TCM constitution categories corresponding to the carotid artery plaque patient samples, and p i is the proportion of samples belonging to category i to all samples.

[0093] The information entropy obtained by dividing the data set D according to attribute A is calculated by formula (3);

[0094]

[0095] In the formula, Info A(D) is the information entropy obtained after dividing the data set D according to attribute A, k is the number of possible values ​​of attribute A, D j is the subset corresponding to the jth value of attribute A, |D j | is a subset D j The number of samples in dataset D, |D| is the total number of samples in dataset D.

[0096] After the decision tree model is established, it is verified through the test set and evaluated through evaluation indicators, including classification accuracy, recall rate, false alarm rate and precision.

[0097] S108, obtaining facial video data of the face to be tested and basic information data to be tested of the patient with carotid artery plaque to be tested, extracting characteristic parameters of the heart rate variability to be tested and basic information features to be tested corresponding to the facial video data of the face to be tested and the basic information data to be tested, inputting the characteristic parameters of the heart rate variability to be tested and the basic information features to be tested into the optimal decision tree model to obtain a result of identifying the TCM constitution type corresponding to the patient with carotid artery plaque to be tested.

[0098] Figure 2 Schematic diagram of the main components of the TCM constitution intelligent classification system for carotid artery plaque patients in an embodiment of the present invention. Figure 2 As shown, the TCM constitution intelligent typing system 1 for patients with carotid plaque provided by an embodiment of the present invention includes an acquisition module 10, a first extraction module 20, a second extraction module 30, a third extraction module 40, a survey module 50, a training set construction module 60, a model construction module 70 and a result acquisition module 80.

[0099] An acquisition module 10 is used to acquire facial video data and basic information data of patients with carotid artery plaques;

[0100] A first extraction module 20 is used to extract basic information features corresponding to the basic information data;

[0101] A second extraction module 30 is used to extract the photoplethysmography signal corresponding to the facial video data;

[0102] A third extraction module 40 is used to extract the heart rate variability characteristic parameters corresponding to the photoplethysmography signal;

[0103] A survey module 50 is configured to respond to a request for obtaining a constitution type of a carotid artery plaque patient and conduct a questionnaire survey on the carotid artery plaque patient to obtain a TCM constitution type identification result corresponding to the carotid artery plaque patient;

[0104] A training set construction module 60 is used to associate the heart rate variability characteristic parameters, basic information characteristics and TCM constitution type identification results to construct a training data set;

[0105] A model building module 70 is used to build a decision tree model and train the decision tree model based on the training data set to obtain an optimal decision tree model;

[0106] The result acquisition module 80 is used to obtain facial video data of the face to be tested and basic information data to be tested of the patient with carotid artery plaque to be tested, extract the heart rate variability characteristic parameters to be tested and the basic information characteristics to be tested corresponding to the facial video data to be tested and the basic information data to be tested, and input the heart rate variability characteristic parameters to be tested and the basic information characteristics to be tested into the optimal decision tree model to obtain the TCM constitution type identification result corresponding to the patient with carotid artery plaque to be tested.

[0107] Figure 3 A schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention, such as Figure 3 As shown, the electronic device 90 includes: a processor 901 (processor), a memory 902 (memory) and a bus 903;

[0108] The processor 901 and the memory 902 communicate with each other via the bus 903.

[0109] The processor 901 is used to call the program instructions in the memory 902 to execute the methods provided by the above-mentioned method embodiments, so as to execute the methods provided by the implementation methods of the present invention.

[0110] This embodiment provides a non-transitory computer-readable storage medium, which stores computer instructions. The computer instructions enable a computer to execute the method provided by the embodiment of the present invention.

[0111] Those skilled in the art will understand that all or part of the steps of implementing the above-mentioned method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above-mentioned method embodiment; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk, etc. Various storage media that can store program codes.

[0112] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A method for intelligent TCM constitution typing of patients with carotid artery plaque, characterized in that: include: Collect facial video data and basic information data of patients with carotid artery plaque; Extracting basic information features corresponding to the basic information data; extracting a photoplethysmography signal corresponding to the facial video data; Extracting heart rate variability characteristic parameters corresponding to the photoplethysmography signal; In response to a request to obtain a constitution type of a patient with carotid artery plaque, conducting a questionnaire survey on the patient with carotid artery plaque to obtain a result of identifying a TCM constitution type corresponding to the patient with carotid artery plaque; Correlate the heart rate variability characteristic parameters, basic information characteristics and TCM constitution type identification results to construct a training data set; Constructing a decision tree model, and training the decision tree model based on the training data set to obtain an optimal decision tree model; Acquire facial video data of the face to be tested and basic information data to be tested of the patient with carotid artery plaque to be tested, extract characteristic parameters of heart rate variability to be tested and basic information features to be tested corresponding to the facial video data to be tested and the basic information data to be tested, and input the characteristic parameters of heart rate variability to be tested and the basic information features to be tested into the optimal decision tree model to obtain a result of identifying the TCM constitution type corresponding to the patient with carotid artery plaque to be tested.

2. The TCM intelligent classification method for carotid artery plaque patients according to claim 1, characterized in that: The step of extracting a photoplethysmogram signal corresponding to the facial video data includes: Obtaining facial single-channel pixels corresponding to the facial video data of a human face, and obtaining a light change curve corresponding to the facial video data of the human face based on the facial single-channel pixels; A fast independent component analysis algorithm is used to obtain a source signal with the highest Pearson coefficient with the green channel, and the source signal is used as a photoplethysmography signal related to the heart beat; The photoplethysmography signal is processed by wavelet transform, fast Fourier transform and narrowband filtering to obtain a pure photoplethysmography signal.

3. The TCM intelligent classification method for carotid artery plaque patients according to claim 2, characterized in that: The step of extracting the heart rate variability characteristic parameter corresponding to the photoplethysmography signal includes: After the pure photoplethysmography signal is up-sampled by cubic spline interpolation, the peak points are extracted and the time intervals between the peak points are calculated to obtain the RR interval series. Acquire an HRV time domain curve based on the RR interval sequence, and extract a time domain characteristic parameter corresponding to the photoplethysmography signal based on the HRV time domain curve; The Welch power spectrum is used to extract the frequency domain characteristic parameters corresponding to the photoplethysmography signal; The nonlinear characteristic parameters corresponding to the photoplethysmography signal are extracted based on the Poincare scatter plot analysis method; A heart rate variability characteristic parameter is obtained based on the time domain characteristic parameter, the frequency domain characteristic parameter and the nonlinear characteristic parameter.

4. The TCM intelligent classification method for carotid artery plaque patients according to claim 1, characterized in that: The TCM constitution type identification results include balanced constitution, qi deficiency constitution, yang deficiency constitution, yin deficiency constitution, blood stasis constitution, phlegm-damp constitution, damp-heat constitution, qi stagnation constitution and special constitution.

5. The TCM intelligent classification method for carotid artery plaque patients according to claim 1, characterized in that: The constructing of the decision tree model, training the decision tree model based on the training data set to obtain an optimal decision tree model, includes: Calculate the loss function value by formula (1); Gain(A)=Info(D)-Info A (D) Formula (1); Where Gain(A) is the information gain value, INfo(D) is the information entropy of the data set D, INfo A (D) is the information entropy obtained after dividing the data set D according to attribute A; The attribute feature with the largest information gain value Gain(A) is used as the root node, and the child nodes are split from the root node. The information gain is calculated and the child node with the largest information gain is used as the child node until all attribute features are less than the set threshold or no attribute feature is selected.

6. The TCM intelligent classification method for carotid artery plaque patients according to claim 5, characterized in that: The calculation of the loss function value by formula (1) also includes: Calculate the information entropy of data set D using formula (2); Where Info(D) is the information entropy of the dataset D, c is the number of nine TCM constitution categories corresponding to the carotid artery plaque patient samples, and p i is the proportion of samples belonging to category i to all samples.

7. The TCM intelligent classification method for carotid artery plaque patients according to claim 5, characterized in that: The calculation of the loss function value by formula (1) also includes: The information entropy obtained by dividing the data set D according to attribute A is calculated by formula (3); In the formula, Info A (D) is the information entropy obtained after dividing the data set D according to attribute A, k is the number of possible values ​​of attribute A, D j is the subset corresponding to the jth value of attribute A, |D j | is a subset D j The number of samples in dataset D, |D| is the total number of samples in dataset D.

8. A system for intelligent TCM constitution classification of patients with carotid artery plaque, characterized by: include: An acquisition module, used to collect facial video data and basic information data of patients with carotid artery plaque; A first extraction module is used to extract basic information features corresponding to the basic information data; a second extraction module, configured to extract a photoplethysmography signal corresponding to the facial video data; a third extraction module, configured to extract characteristic parameters of heart rate variability corresponding to the photoplethysmography signal; a survey module, configured to respond to a request for obtaining a constitution type of a carotid artery plaque patient and conduct a questionnaire survey on the carotid artery plaque patient to obtain a result of identifying a TCM constitution type corresponding to the carotid artery plaque patient; A training set construction module is used to associate heart rate variability characteristic parameters, basic information characteristics and TCM constitution type identification results to construct a training data set; A model building module is used to build a decision tree model and train the decision tree model based on the training data set to obtain an optimal decision tree model; The result acquisition module is used to obtain facial video data of the face to be tested and basic information data to be tested of the patient with carotid artery plaque to be tested, extract the heart rate variability characteristic parameters to be tested and the basic information features to be tested corresponding to the facial video data to be tested and the basic information data to be tested, and input the heart rate variability characteristic parameters to be tested and the basic information features to be tested into the optimal decision tree model to obtain the TCM constitution type identification result corresponding to the patient with carotid artery plaque to be tested.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A non-transitory computer readable medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images