WO2017016086A1 - 基于生理信息的抑郁症评估系统及其评估方法 - Google Patents

基于生理信息的抑郁症评估系统及其评估方法 Download PDF

Info

Publication number
WO2017016086A1
WO2017016086A1 PCT/CN2015/093158 CN2015093158W WO2017016086A1 WO 2017016086 A1 WO2017016086 A1 WO 2017016086A1 CN 2015093158 W CN2015093158 W CN 2015093158W WO 2017016086 A1 WO2017016086 A1 WO 2017016086A1
Authority
WO
WIPO (PCT)
Prior art keywords
signal
processing
wave
parameter
parameters
Prior art date
Application number
PCT/CN2015/093158
Other languages
English (en)
French (fr)
Inventor
杨荣骞
陈秀文
吕瑞雪
宋传旭
Original Assignee
华南理工大学
深圳市是源医学科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华南理工大学, 深圳市是源医学科技有限公司 filed Critical 华南理工大学
Priority to US15/109,815 priority Critical patent/US20170238858A1/en
Publication of WO2017016086A1 publication Critical patent/WO2017016086A1/zh

Links

Images

Classifications

    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/16Devices for psychotechnics; Testing reaction times ; Devices for evaluating the psychological state
    • A61B5/165Evaluating the state of mind, e.g. depression, anxiety
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/0002Remote monitoring of patients using telemetry, e.g. transmission of vital signals via a communication network
    • A61B5/0015Remote monitoring of patients using telemetry, e.g. transmission of vital signals via a communication network characterised by features of the telemetry system
    • A61B5/0022Monitoring a patient using a global network, e.g. telephone networks, internet
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/01Measuring temperature of body parts ; Diagnostic temperature sensing, e.g. for malignant or inflamed tissue
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/02Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
    • A61B5/0205Simultaneously evaluating both cardiovascular conditions and different types of body conditions, e.g. heart and respiratory condition
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/02Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
    • A61B5/024Detecting, measuring or recording pulse rate or heart rate
    • A61B5/02416Detecting, measuring or recording pulse rate or heart rate using photoplethysmograph signals, e.g. generated by infrared radiation
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/24Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316Modalities, i.e. specific diagnostic methods
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/24Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316Modalities, i.e. specific diagnostic methods
    • A61B5/318Heart-related electrical modalities, e.g. electrocardiography [ECG]
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/24Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316Modalities, i.e. specific diagnostic methods
    • A61B5/369Electroencephalography [EEG]
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/24Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316Modalities, i.e. specific diagnostic methods
    • A61B5/389Electromyography [EMG]
    • A61B5/392Detecting gastrointestinal contractions
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/24Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316Modalities, i.e. specific diagnostic methods
    • A61B5/398Electrooculography [EOG], e.g. detecting nystagmus; Electroretinography [ERG]
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/72Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235Details of waveform analysis
    • A61B5/7264Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
    • A61B5/7267Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16HHEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H40/00ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
    • G16H40/60ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices
    • G16H40/67ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices for remote operation
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/05Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves 
    • A61B5/053Measuring electrical impedance or conductance of a portion of the body
    • A61B5/0531Measuring skin impedance
    • A61B5/0533Measuring galvanic skin response
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/24Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316Modalities, i.e. specific diagnostic methods
    • A61B5/318Heart-related electrical modalities, e.g. electrocardiography [ECG]
    • A61B5/346Analysis of electrocardiograms
    • A61B5/349Detecting specific parameters of the electrocardiograph cycle
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/24Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316Modalities, i.e. specific diagnostic methods
    • A61B5/389Electromyography [EMG]
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/48Other medical applications
    • A61B5/4806Sleep evaluation
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/72Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235Details of waveform analysis
    • A61B5/7253Details of waveform analysis characterised by using transforms
    • A61B5/7257Details of waveform analysis characterised by using transforms using Fourier transforms
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/72Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235Details of waveform analysis
    • A61B5/7253Details of waveform analysis characterised by using transforms
    • A61B5/726Details of waveform analysis characterised by using transforms using Wavelet transforms
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16HHEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/70ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients

Definitions

  • the invention relates to a depression evaluation technology, in particular to a physiological information-based depression evaluation system and an evaluation method thereof.
  • the pathogenesis of depression is mostly concentrated on neurotransmitters and their receptors, especially monoamine neurotransmitters and their receptors. It is believed that neuropeptides play an important role in the pathogenesis of depression. But so far, there is no unified conclusion about the pathogenesis of depression.
  • physiological information such as ECG, pulse wave, EEG, skin electricity, stomach power, myoelectricity, ocular electricity, polysomnography, and temperature in patients with depression are different from those in normal people.
  • the time domain, frequency domain, and time domain geometric parameters of the electrical signal are different. Therefore, according to the differences in the performance of various physiological information, the signal is processed, a large number of signal parameters are calculated, and the mathematical model of depression assessment is established to evaluate depression, which has research basis, feasibility and clinical applicability.
  • the primary object of the present invention is to overcome the shortcomings and deficiencies of the existing depression evaluation techniques and to provide a A physiological information-based depression assessment system that collects human ECG information and one or more physiological information of pulse wave, EEG, galvanic, gastric, electromyographic, ocular, polysomn, and temperature. Calculate the time domain and frequency domain parameters of physiological information, extract feature parameter sets, establish a mathematical model of depression assessment, and then evaluate whether the subject has depression and depression level.
  • Another object of the present invention is to overcome the shortcomings and deficiencies of the existing depression evaluation method, and to provide an evaluation method applied to a physiological information-based depression evaluation system, which can objectively and quantitatively assess whether a subject suffers from Depression and depression levels.
  • a physiological information-based depression evaluation system comprising: an information acquisition module, a signal processing module, a parameter calculation module, a feature selection module, a machine learning module, and an output result module.
  • the information collecting module is configured to collect an ECG signal and selectively acquire one of a pulse wave signal, an EEG signal, a skin electrical signal, an gastric electrical signal, an EMG signal, an EOG signal, a polysomnographic signal, and a temperature signal. More than one type of physiological information.
  • the signal collected by the information acquisition module is transmitted to the signal processing module by means of USB serial cable transmission or Bluetooth wireless transmission.
  • a signal processing module configured to perform signal processing on physiological information, including an electrocardiographic signal processing unit, a pulse wave signal processing unit, an electroencephalogram signal processing unit, a skin electrical signal processing unit, a gastric electrical signal processing unit, an electromyography signal processing unit, An ocular signal processing unit, a polysomnographic signal processing unit, and a temperature signal processing unit.
  • the central electrical signal processing unit includes de-baseline processing, filter denoising processing, extraction of sinus beat interval (RR interval) processing, interpolation processing, Fourier transform processing, and spectral analysis and spectral estimation processing.
  • the pulse wave signal processing unit includes de-baseline processing, filter denoising processing, extraction pulse interval (PP interval) processing, interpolation processing, Fourier transform processing, and spectral analysis and spectral estimation processing.
  • the EEG signal processing unit includes de-baseline processing, threshold denoising processing, wavelet decomposition processing, and spectral analysis and spectral estimation processing.
  • the electrical electrical signal processing unit includes a de-baseline processing and a wavelet filtering process.
  • the gastric electrical signal processing unit includes de-baseline processing, Hilbert-Huang transform processing, wavelet analysis processing, multi-resolution analysis processing, and independent component analysis processing.
  • the EMG signal processing unit includes de-baseline processing and wavelet packet adaptive threshold denoising.
  • the EO signal processing unit includes de-baseline processing, weighted median filtering processing, and wavelet transform processing.
  • the polysomnographic signal processing unit includes processing a sleep brain electrical signal, a sleep myoelectric signal, and a sleep ocular electrical signal, performing de-baseline processing, threshold denoising processing, wavelet decomposition processing, spectrum analysis, and spectral estimation processing on the sleep brain electrical signal.
  • De-baseline processing, weighted median filtering processing, and wavelet transform processing are performed on the sleep electro-oculogram signal, and the sleep electromyogram signal is subjected to de-baseline processing, wavelet packet adaptive threshold denoising processing, and sleep staging processing.
  • Temperature signal processing unit includes de-baseline processing, threshold filtering processing, and construction The relationship between the temperature value and the gray value of the image.
  • the signal processing module outputs the processed signal to the parameter calculation module.
  • the parameter calculation module is configured to calculate signal parameters of the processed signal, including an electrocardiogram parameter calculation unit, a pulse wave parameter calculation unit, an electroencephalogram parameter calculation unit, a skin electrical parameter calculation unit, a gastric electrical parameter calculation unit, and an electromyogram parameter calculation. Unit, electrooculogram parameter calculation unit, polysomnography parameter calculation unit, and temperature parameter calculation unit.
  • the central electrical parameter calculation unit includes calculating the RR interval, the mean of all RR intervals (Mean), the standard deviation of the heartbeat interval (SDNN), the root mean square (RMSSD) of the difference between adjacent heartbeat intervals, and a 50-millisecond interval.
  • the pulse wave parameter calculation unit includes calculating the PP interval, the mean of all PP intervals (Mean), the standard deviation of the pulse interval (SDNN), the root mean square (RMSSD) of the difference between adjacent pulse intervals, and the interval of 50 msec or more.
  • Proportion of adjacent pulse interval differences PNN50
  • standard deviation between adjacent pulse intervals SDSD
  • VLF very low frequency components
  • LF low frequency components
  • HF high frequency components
  • TP total spectrum power
  • SD2 slope of short-term detrended fluctuation analysis
  • a2 slope of long-term de-trend fluctuation analysis
  • the EEG parameter calculation unit includes calculation of delta wave amplitude, ⁇ wave power, ⁇ wave mean, ⁇ wave variance, ⁇ wave hemiplegia, ⁇ wave kurtosis, ⁇ wave amplitude, ⁇ wave power, ⁇ wave mean, ⁇ wave Variance, ⁇ wave hemiplegia, ⁇ wave kurtosis, ⁇ wave amplitude, ⁇ wave power, ⁇ wave mean, ⁇ wave variance, ⁇ wave hemiplegia, ⁇ wave kurtosis, ⁇ wave amplitude, ⁇ wave power, ⁇ wave mean, ⁇ wave variance, ⁇ wave hemiplegia, ⁇ wave kurtosis and wavelet entropy.
  • the skin electrical parameter calculation unit includes calculating a skin sympathetic response latency, a skin sympathetic response amplitude, and a skin resistance value.
  • the gastric electrical parameter calculation unit includes calculating a normal gastric electrical rhythm, a slow wave, a hyperkinetic component, and a tachycardia component.
  • the myoelectric parameter calculation unit includes a calculation base value, a minimum value, a maximum value, a myoelectric decline ability, and an electromyogram curve.
  • the electrooculogram parameter calculation unit includes calculating an R wave component, an r wave component, an S wave component, and an s wave component.
  • the polysomnographic signal parameter calculation unit includes calculating a sleep latency, a total sleep time, an arousal index, a sleep period (S1), a shallow sleep period (S2), a moderate sleep period (S3), a deep sleep period (S4), and rapid eye movement. Percentage, number of rapid eye movement sleep cycles, rapid eye movement sleep latency, rapid eye movement sleep intensity, rapid eye movement sleep density, and rapid eye movement sleep time.
  • the temperature parameter calculation unit includes calculating a temperature distribution in the human body.
  • the parameter calculation module outputs signal parameters to the feature selection module.
  • a feature selection module is configured to obtain a feature parameter set related to a depression level among all signal parameters.
  • the feature selection module outputs a feature parameter set to the machine learning module.
  • the machine learning module is used to train the classifier of the depression level quantification, and the feature parameter set is used to establish a mathematical model of depression assessment to quantify the level of depression.
  • the machine learning module outputs a depression level to the output result module.
  • An output result module is used to display the level of depression output by the mathematical model of the depression assessment.
  • an evaluation method applied to a physiological information-based depression evaluation system which may include the following steps:
  • Step 1 Signal processing of the ECG signal and simultaneously signal one of a pulse wave signal, an EEG signal, a skin electrical signal, an gastric electrical signal, an EMG signal, an EEG signal, a polysomn signal, and a temperature signal or More than one signal is used for signal processing and the signal parameters of the processed signal are calculated. among them:
  • Skin electrical signal processing and parameter calculation The skin sympathetic response latency, skin sympathetic response amplitude and skin resistance value were calculated by skin electrical signal to baseline processing and wavelet filtering;
  • Gastric electrical signal processing and parameter calculations calculate normal gastric electrical rhythm, slow wave, gastric hyperactivity and gastric motility through gastric electrical signal de-baseline processing, Hilbert-Huang transform processing, wavelet analysis processing, multi-resolution analysis processing, and independent component analysis processing. Overspeed component
  • EMG signal processing and parameter calculation The baseline value, minimum value, highest value, myoelectric decline ability and myoelectric curve were calculated by the EMG signal to baseline processing and the wavelet packet adaptive threshold denoising process;
  • Electro-oculogram signal processing and parameter calculation The R wave component, the r wave component, the S wave component and the s wave component are calculated by the EOG de-baseline processing, the weighted median filtering process, and the wavelet transform process;
  • Temperature signal processing and parameter calculation The temperature distribution in the human body is calculated by the temperature signal to baseline processing, the threshold filtering process, and the relationship between the temperature value and the gray value of the image.
  • Step 2 normalize the signal parameters calculated in step 1, and perform feature selection on the parameter set composed of the normalized signal parameters to obtain a feature parameter set.
  • the normalized processing method :
  • X is the signal parameter of the parameter set
  • X i is the i-th normalized signal parameter value
  • X in is the i-th normalized value
  • X imean is the normal mean of the i-th parameter.
  • X istd represents the normal standard deviation of the ith parameter, and i is a positive integer.
  • the feature selection is divided into two parts: feature search and evaluation criteria, wherein the search algorithm uses one or more combinations of the following algorithms: Complete Search, Sequential Search, Random Search Algorithm (Random) Search), Genetic Algorithm, Simulated Annealing, traceable greedy search expansion algorithm, evaluation criteria can optionally use Wapper model or CfsSubsetEval attribute evaluation method.
  • the ECG and pulse wave signals are acquired during the evaluation process.
  • the feature selection is combined with the full search algorithm and the Wapper model. During the evaluation process, ECG, EKG and polysomnography signals are acquired. The feature selection is combined with a random search algorithm. The way the CfsSubsetEval property evaluates methods. According to different types of acquired signals, select a combination of algorithms with appropriate and high accuracy.
  • Step 3 Perform machine learning according to the feature parameter set obtained in step 2, and establish a mathematical model of depression assessment in the process of machine learning using the feature parameter set.
  • the machine learning algorithm may selectively use one or more of the following algorithms: Bayesian, Decision Tree, AdaBoost, k-Nearest Neighbor ), Support Vector Machine (SVM).
  • SVM Support Vector Machine
  • Y is the output value of the mathematical model of depression assessment
  • n is the number of machine learning algorithms selected for use
  • y i is the output value of the i-th algorithm
  • a i is the coefficient of the i-th algorithm
  • i is a positive integer.
  • the present invention has the following advantages and beneficial effects:
  • the present invention relates to an electrocardiographic signal and a signal or more than one of a pulse wave signal, an EEG signal, a skin electrical signal, an gastric electrical signal, an EMG signal, an EOG signal, a polysomnographic signal, and a temperature signal.
  • Signal combination, signal processing, parameter calculation, mathematical model establishment, multiple signal combinations can be selected for evaluation, flexibility and novelty;
  • the present invention proposes a method for normalizing signal parameters, comparing parameters with mean and standard deviations in normal samples, eliminating differences in numerical values and deviations of parameters, and making parameter set feature selection more scientific and effective;
  • the present invention proposes a combination of multiple feature selection and machine learning algorithms, and the mathematical model is more flexible in terms of signal types;
  • Figure 1 is a schematic diagram of a depression assessment system based on physiological information.
  • FIG. 2 is a structural diagram of a depression information evaluation system based on physiological information.
  • a physiological information-based depression evaluation system includes: an information acquisition module, a signal processing module, a parameter calculation module, a feature selection module, a machine learning module, and an output result module; and the signal collected by the information acquisition module passes USB serial cable transmission or Bluetooth wireless transmission is transmitted to the signal processing module.
  • the signal processing module outputs the processed signal to the parameter calculation module.
  • the parameter calculation module outputs signal parameters to the feature selection module.
  • the feature selection module outputs a feature parameter set to the machine learning module.
  • the machine learning module outputs a depression level to the output result module.
  • the structure of the physiological information-based depression evaluation system is as shown in FIG. 2, and the information collection module is configured to collect an electrocardiogram signal and collect a pulse wave signal, an EEG signal, a skin electrical signal, a gastric electrical signal, and a muscle.
  • the signal processing module is configured to process physiological information, including de-baseline processing, filtering denoising processing, extracting heartbeat interval processing, time-frequency transform processing, and spectrum analysis and spectrum estimation processing.
  • the parameter calculation module is configured to calculate a signal parameter of the processed signal, including a time domain parameter of a heart rate variability, a frequency domain parameter, and a time domain geometric parameter, and selectively calculate a pulse wave signal according to the collected physiological information, Time domain parameters, frequency domain parameters, histogram parameters, distribution maps of one or more signals in brain electrical signals, electrical signals, gastric electrical signals, myoelectric signals, ocular electrical signals, polysomnographic signals, and temperature signals parameter.
  • the feature selection module is configured to obtain a feature parameter set related to a depression level among all signal parameters.
  • the machine learning module is configured to train a classifier for quantifying the level of depression, and use the feature parameter set to establish a mathematical model of depression assessment to quantify the level of depression.
  • the output result module is configured to display a depression level output by a mathematical model of depression assessment.
  • Step 1 Obtain physiological information, including electrocardiogram, and one or more physiological information of pulse wave, brain electricity, skin electricity, stomach electricity, myoelectricity, ocular electricity, polysomnography, and temperature. among them:
  • the ECG signal acquisition can be selected to measure the ECG signal in a resting state of five minutes, and the sampling rate of the ECG acquisition can be selected at 500 Hz or more;
  • Pulse wave acquisition can selectively use the pulse sensor of the infrared light transmission tip to output the blood volume change of the end of the blood vessel to collect the pulse signal, or use the shock measurement method to collect the wrist pulse signal.
  • Pulse wave acquisition sampling rate can be selected 500Hz or more;
  • EEG acquisition can choose to use 10-20 system point excitation to collect spontaneous brain electrical activity in the cerebral cortex;
  • the skin electrical collection was tested by skin sympathetic response, single pulse percutaneous electrical stimulation of the median nerve of the wrist, testing the initial latency and amplitude of the skin sympathetic response, and testing the skin resistance values of the right hand large fish muscle and forearm volar side;
  • Gastric electricity collection uses gastric surface electrodes placed in the upper abdomen to measure gastric myoelectric activity
  • the myoelectric collection is stimulated by a biofeedback device, and the myoelectric electrode connected to the forehead is used to measure the signal of the myoelectricity;
  • Polysomnography measures sleep time and its parameters by simultaneously collecting ocular electricity, mandibular electromyography and EEG;
  • Temperature acquisition can use the infrared temperature measurement principle to measure the temperature of the body.
  • Signal acquisition is a common signal acquisition.
  • Step 2 Perform signal processing on the physiological information acquired in step 1, and calculate signal parameters; the specific parameter list is shown in Table 1 below, and Table 1 is a list of electrical signals and their parameter descriptions:
  • ECG signal processing and parameter calculation calculate RR interval, Mean, through ECG signal to baseline processing, filter denoising processing, extraction RR interval processing, interpolation processing, Fourier transform processing, and spectral analysis and spectral estimation processing.
  • Skin electrical signal processing and parameter calculation The skin sympathetic response latency, skin sympathetic response amplitude and skin resistance value were calculated by skin electrical signal to baseline processing and wavelet filtering;
  • Gastric electrical signal processing and parameter calculations calculate normal gastric electrical rhythm, slow wave, gastric hyperactivity and gastric motility through gastric electrical signal de-baseline processing, Hilbert-Huang transform processing, wavelet analysis processing, multi-resolution analysis processing, and independent component analysis processing. Overspeed component
  • EMG signal processing and parameter calculation The baseline value, minimum value, highest value, myoelectric decline ability and myoelectric curve were calculated by the EMG signal to baseline processing and the wavelet packet adaptive threshold denoising process;
  • Electro-oculogram signal processing and parameter calculation The R wave component, the r wave component, the S wave component and the s wave component are calculated by the EOG de-baseline processing, the weighted median filtering process, and the wavelet transform process;
  • Temperature signal processing and parameter calculation The temperature distribution in the human body is calculated by the temperature signal to baseline processing, the threshold filtering process, and the relationship between the temperature value and the gray value of the image.
  • Step 3 normalize the signal parameters calculated in step 2, perform feature selection on the parameter set composed of the normalized signal parameters, and obtain a feature parameter set, and the normalization processing method is as follows:
  • X is the signal parameter of the parameter set
  • X i is the i-th normalized signal parameter value
  • X in is the i-th normalized value
  • X imean is the normal mean of the i-th parameter.
  • X istd represents the normal standard deviation of the ith parameter, and i is a positive integer.
  • the feature selection is divided into two parts: feature search and evaluation criteria, wherein the search algorithm uses one or more combinations of the following algorithms: Complete Search, Sequential Search, Random Search Algorithm (Random) Search), Genetic Algorithm, Simulated Annealing, traceable greedy search expansion algorithm, evaluation criteria can optionally use Wapper model or CfsSubsetEval attribute evaluation method.
  • the ECG and pulse wave signals are acquired during the evaluation process.
  • the feature selection is combined with the full search algorithm and the Wapper model. During the evaluation process, ECG, EKG and polysomnography signals are acquired. The feature selection is combined with a random search algorithm. The way the CfsSubsetEval property evaluates methods. According to different types of acquired signals, select a combination of algorithms with appropriate and high accuracy.
  • Step 4 Perform machine learning according to the feature parameter set obtained in step 3, and establish a mathematical model of depression assessment in the process of machine learning using the feature parameter set.
  • the machine learning algorithm may selectively use one or more of the following algorithms: Bayesian, Decision Tree, AdaBoost, k-Nearest Neighbor ), Support Vector Machine (SVM).
  • SVM Support Vector Machine
  • Y is the output value of the mathematical model of depression assessment
  • n is the number of machine learning algorithms selected for use
  • y i is the output value of the i-th algorithm
  • a i is the coefficient of the i-th algorithm
  • i is a positive integer.
  • the mathematical model of depression assessment establishes a mathematical model of depression assessment based on a variety of physiological information, and evaluates the level of depression using the output of the mathematical model of depression assessment, and classifies the depression into five levels: normal, general, mild depression, Moderate depression and major depression.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Biomedical Technology (AREA)
  • Medical Informatics (AREA)
  • Public Health (AREA)
  • General Health & Medical Sciences (AREA)
  • Surgery (AREA)
  • Molecular Biology (AREA)
  • Animal Behavior & Ethology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Pathology (AREA)
  • Veterinary Medicine (AREA)
  • Biophysics (AREA)
  • Cardiology (AREA)
  • Psychiatry (AREA)
  • Physiology (AREA)
  • Artificial Intelligence (AREA)
  • Psychology (AREA)
  • Developmental Disabilities (AREA)
  • Child & Adolescent Psychology (AREA)
  • Social Psychology (AREA)
  • Hospice & Palliative Care (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Educational Technology (AREA)
  • Pulmonology (AREA)
  • Fuzzy Systems (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Signal Processing (AREA)
  • Evolutionary Computation (AREA)
  • Mathematical Physics (AREA)
  • Gastroenterology & Hepatology (AREA)
  • Ophthalmology & Optometry (AREA)
  • Business, Economics & Management (AREA)
  • General Business, Economics & Management (AREA)
  • Epidemiology (AREA)
  • Primary Health Care (AREA)
  • Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)

Abstract

本发明公开了一种基于生理信息的抑郁症评估系统,包括:信息采集模块、信号处理模块、参数计算模块、特征选择模块、机器学习模块和输出结果模块。本发明还公开了一种基于多种生理信息的抑郁症评估方法,包括以下步骤:1、对心电信号以及脉搏波信号、脑电信号、皮电信号、胃电信号、肌电信号、眼电信号、多导睡眠信号和温度信号中一种或一种以上信号进行信号处理,并计算信号参数;2、利用得到的信号参数进行归一化处理,对经过归一化处理的信号参数组成的参数集进行特征选择,得到特征参数集;3、利用得到的特征参数集进行机器学习,利用特征参数集与抑郁等级的关系建立抑郁评估数学模型评估抑郁等级。具有能避免量表评估的主观性等优点。

Description

基于生理信息的抑郁症评估系统及其评估方法 技术领域
本发明涉及一种抑郁症评估技术,特别涉及一种基于生理信息的抑郁症评估系统及其评估方法。
背景技术
随着社会发展,人们面临越来越多的压力,抑郁症的发病率也越来越高。根据调查中国约有9000万抑郁症患者,占总人口的6.4%。全世界抑郁症患者约有3.5亿。抑郁症患者一般表现为情绪低沉,对以前感兴趣的事物丧失兴趣以及注意力降低。抑郁症等级有轻度、中度、重度的区别,患病情况严重的有自杀倾向。抑郁症的病因是复杂的,而不是单一的,主要由生物、心理和社会因素共同组成生物-心理-社会的统一模式,有遗传因素、生物化学因素、神经内分泌因素、心理社会因素等原因影响。抑郁症的发病机制研究多集中于神经递质及其受体,尤其是单胺类神经递质及其受体,研究认为神经肽在抑郁症发病中起重要作用。但是至今,抑郁症的发病机制还没有一个统一的定论。
如今临床上对抑郁症的评估主要根据病史、临床症状等方式,目前国际上通用的评估标准有ICD-10和DSM-IV。国内主要采用ICD-10,通过抑郁症症状的表现以及抑郁症自评量表(SDS)判断受测者是否患有抑郁症,这样的评估方式会受到受测者主观陈述、心理医生的自身主观因素和临床经验的影响,并不是客观评估抑郁症的有效方法。因此需要一种基于生理信息对抑郁症进行评估,客观量化是否患有抑郁症以及抑郁等级。
根据以往的研究,抑郁症患者的心电、脉搏波、脑电、皮电、胃电、肌电、眼电、多导睡眠、温度等生理信息跟正常人有所差异。表现为电信号的时域、频域、时域几何参数等有所不同。因此根据多种生理信息表现的差异,对信号进行处理,计算大量的信号参数,建立抑郁评估数学模型评估抑郁症具有研究基础、可行性和临床实用性。
发明内容
本发明的首要目的在于克服现有抑郁症评价技术的缺点和不足,提供一种 基于生理信息的抑郁症评估系统,该系统通过采集人体心电信息以及脉搏波、脑电、皮电、胃电、肌电、眼电、多导睡眠、温度中一种或一种以上生理信息,计算生理信息的时域、频域等参数,提取特征参数集,建立抑郁评估数学模型,进而对受测者是否患有抑郁症以及抑郁等级进行评估。
本发明的另一目的在于克服现有抑郁症评价方法的缺点和不足,提供一种应用于基于生理信息的抑郁症评估系统的评估方法,该评估方法能够客观量化地评估受测者是否患有抑郁症以及抑郁等级。
本发明的首要目的通过下述技术方案实现:一种基于生理信息的抑郁症评估系统,包括:信息采集模块、信号处理模块、参数计算模块、特征选择模块、机器学习模块和输出结果模块。
信息采集模块,用于采集心电信号以及选择性地采集脉搏波信号、脑电信号、皮电信号、胃电信号、肌电信号、眼电信号、多导睡眠信号、温度信号中一种或一种以上的生理信息。信息采集模块采集的信号通过USB串口有线传输或者蓝牙无线传输的方式传输到信号处理模块中。
信号处理模块,用于对生理信息进行信号处理,包括心电信号处理单元、脉搏波信号处理单元、脑电信号处理单元、皮电信号处理单元、胃电信号处理单元、肌电信号处理单元、眼电信号处理单元、多导睡眠信号处理单元和温度信号处理单元。其中心电信号处理单元包括去基线处理、滤波去噪处理、提取窦性心搏间期(RR间期)处理、插值处理、傅里叶变换处理以及谱分析和谱估计处理。脉搏波信号处理单元包括去基线处理、滤波去噪处理、提取脉搏间期(PP间期)处理、插值处理、傅里叶变换处理以及谱分析和谱估计处理。脑电信号处理单元包括去基线处理、阈值去噪处理、小波分解处理以及谱分析和谱估计处理。皮电信号处理单元包括去基线处理和小波滤波处理。胃电信号处理单元包括去基线处理、Hilbert-Huang变换处理、小波分析处理、多分辨率分析处理和独立成分分析处理。肌电信号处理单元包括去基线处理和小波包自适应阈值去噪。眼电信号处理单元包括去基线处理、加权中值滤波处理和小波变换处理。多导睡眠信号处理单元包括处理睡眠脑电信号、睡眠肌电信号和睡眠眼电信号,对所述睡眠脑电信号进行去基线处理、阈值去噪处理、小波分解处理以及谱分析和谱估计处理,对所述睡眠眼电信号进行去基线处理、加权中值滤波处理和小波变换处理,对所述睡眠肌电信号进行去基线处理、小波包自适应阈值去噪处理和睡眠分期处理。温度信号处理单元包括去基线处理、阈值滤波处理、建 立温度值与图像灰度值的关系式。信号处理模块输出经过处理的信号到参数计算模块。
参数计算模块,用于计算经过处理的信号的信号参数,包括心电参数计算单元、脉搏波参数计算单元、脑电参数计算单元、皮电参数计算单元、胃电参数计算单元、肌电参数计算单元、眼电参数计算单元、多导睡眠参数计算单元和温度参数计算单元。其中心电参数计算单元包括计算RR间期、所有RR间期的均值(Mean)、心跳间期的标准差(SDNN)、相邻心跳间期差值的均方根(RMSSD)、50毫秒间隔以上相邻心跳间期差值的比例(PNN50)、相邻心跳间期之间的标准差(SDSD)、极低频成分(VLF)、低频成分(LF)、高频成分(HF)、频谱总功率(TP)、低频成分与高频成分的比值(LF/HF)、RR间期散点图中垂直于y=x的标准偏差(SD1)、RR间期散点图中y=x直线的标准偏差(SD2)、短期去趋势波动分析的斜率(a1)和长期去趋势波动分析的斜率(a2)。脉搏波参数计算单元包括计算PP间期、所有PP间期的均值(Mean)、脉搏间期的标准差(SDNN)、相邻脉搏间期差值的均方根(RMSSD)、50毫秒间隔以上相邻脉搏间期差值的比例(PNN50)、相邻脉搏间期之间的标准差(SDSD)、极低频成分(VLF)、低频成分(LF)、高频成分(HF)、频谱总功率(TP)、低频成分与高频成分的比值(LF/HF)、PP间期散点图中垂直于y=x的标准偏差(SD1)、PP间期散点图中y=x直线的标准偏差(SD2)、短期去趋势波动分析的斜率(a1)和长期去趋势波动分析的斜率(a2)。脑电参数计算单元包括计算δ波幅值、δ波功率、δ波均值、δ波方差、δ波偏歪度、δ波峭度、θ波幅值、θ波功率、θ波均值、θ波方差、θ波偏歪度、θ波峭度、α波幅值、α波功率、α波均值、α波方差、α波偏歪度、α波峭度、β波幅值、β波功率、β波均值、β波方差、β波偏歪度、β波峭度和小波熵。皮电参数计算单元包括计算皮肤交感反应潜伏期、皮肤交感反应波幅和皮肤电阻值。胃电参数计算单元包括计算正常胃电节律、慢波、胃动过缓成分和胃动过速成分。肌电参数计算单元包括计算基础值、最小值、最高值、肌电下降能力和肌电曲线。眼电参数计算单元包括计算R波成分、r波成分、S波成分和s波成分。多导睡眠信号参数计算单元包括计算睡眠潜伏期、睡眠总时间、觉醒指数、入睡期(S1)、浅睡期(S2)、中度睡眠期(S3)、深度睡眠期(S4)、快速眼动百分比、快速眼动睡眠周期数、快速眼动睡眠潜伏期,快速眼动睡眠强度,快速眼动睡眠密度和快速眼动睡眠时间。温度参数计算单元包括计算人体体内温度分布。参数计算模块输出信号参数到特征选择模块。
特征选择模块,用于在全部信号参数中获取与抑郁等级相关的特征参数集。特征选择模块输出特征参数集到机器学习模块。
机器学习模块,用于训练抑郁等级量化的分类器,利用特征参数集建立抑郁评估数学模型,量化抑郁等级。机器学习模块输出抑郁等级到输出结果模块。
输出结果模块,用于显示抑郁评估数学模型输出的抑郁等级。
本发明的另一目的通过下述技术方案实现:一种应用于基于生理信息的抑郁症评估系统的评估方法,可以包括以下步骤:
步骤1:对心电信号进行信号处理并同时对脉搏波信号、脑电信号、皮电信号、胃电信号、肌电信号、眼电信号、多导睡眠信号和温度信号中的一种信号或一种以上的信号进行信号处理,并计算经过处理的信号的信号参数。其中:
心电信号处理和参数计算通过心电信号去基线处理、滤波去噪处理、提取RR间期处理、插值处理、傅里叶变换处理以及谱分析和谱估计处理计算RR间期、Mean、SDNN、RMSSD、PNN50、SDSD VLF、LF、HF、TP、LF/HF、SD1、SD2、a1和a2;
脉搏波信号处理和参数计算通过脉搏波信号去基线处理、滤波去噪处理、提取PP间期处理、插值处理、傅里叶变换处理以及谱分析和谱估计处理;
脑电信号处理和参数计算通过脑电信号去基线处理、阈值去噪处理、小波分解处理以及谱分析和谱估计处理计算δ波幅值、δ波功率、δ波均值、δ波方差、δ波偏歪度、δ波峭度、θ波幅值、θ波功率、θ波均值、θ波方差、θ波偏歪度、θ波峭度、α波幅值、α波功率、α波均值、α波方差、α波偏歪度、α波峭度、β波幅值、β波功率、β波均值、β波方差、β波偏歪度、β波峭度和小波熵;
皮电信号处理和参数计算通过皮电信号去基线处理和小波滤波计算皮肤交感反应潜伏期、皮肤交感反应波幅和皮肤电阻值;
胃电信号处理和参数计算通过胃电信号去基线处理、Hilbert-Huang变换处理、小波分析处理、多分辨率分析处理和独立成分分析处理计算正常胃电节律、慢波、胃动过缓和胃动过速成分;
肌电信号处理和参数计算通过肌电信号去基线处理和小波包自适应阈值去噪处理计算基础值、最小值、最高值、肌电下降能力和肌电曲线;
眼电信号处理和参数计算通过眼电信号去基线处理、加权中值滤波处理和小波变换处理计算R波成分、r波成分、S波成分和s波成分;
多导睡眠信号处理和参数计算通过睡眠脑电信号去基线处理、阈值去噪处理、小波分解处理以及谱分析和谱估计处理,睡眠眼电信号去基线处理、加权中值滤波处理和小波变换处理,睡眠肌电信号去基线处理、小波包自适应阈值去噪处理和睡眠分期处理计算睡眠潜伏期、睡眠总时间、觉醒指数、S1、S2、S3、S4、快速眼动百分比、快速眼动睡眠周期数、快速眼动睡眠潜伏期、快速眼动睡眠强度、快速眼动睡眠密度和快速眼动睡眠时间;
温度信号处理和参数计算通过温度信号去基线处理、阈值滤波处理和建立温度值与图像灰度值的关系式计算人体体内温度分布。
步骤2:利用步骤1计算得到的信号参数进行归一化处理,对经过归一化处理后的信号参数组成的参数集进行特征选择,得到特征参数集。所述的归一化处理方法:
Figure PCTCN2015093158-appb-000001
其中,X是指参数集的信号参数,Xi表示第i个进行归一化的信号参数值,Xin表示第i个归一化后的值,Ximean表示第i个参数的正常均值,Xistd表示第i个参数的正常标准差,i是正整数。所述的特征选择分为特征搜索和评价准则两部分,其中搜索算法使用以下算法中一种或一种以上的组合:完全搜索(Complete Search)、顺序搜索(Sequential Search)、随机搜索算法(Random Search)、遗传算法(Genetic Algorithm)、模拟退火搜索算法(Simulated Annealing)、可回溯的贪婪搜索扩张算法,评价准则可选择性地使用Wapper模型或CfsSubsetEval属性评估方法。其中在评估过程中获取心电和脉搏波信号,特征选择采用结合完全搜索算法与Wapper模型的方式;评估过程中,获取心电、皮电和多导睡眠信号,特征选择采用结合随机搜索算法与CfsSubsetEval属性评估方法的方式。根据采集信号种类不同,选择合适、准确度高的算法组合。
步骤3:根据步骤2得到的特征参数集进行机器学习,使用特征参数集在机器学习的过程中建立抑郁评估数学模型。其中机器学习的算法可选择性地使用以下算法中一种或一种以上组合:贝叶斯分类器(Bayes)、决策树算法(Decision Tree)、AdaBoost算法、k-近邻法(k-Nearest Neighbor)、支持向量机(SVM)。抑郁评估数学模型的表达式为:
Figure PCTCN2015093158-appb-000002
其中,Y是抑郁评估数学模型输出值,n是选择使用的机器学习算法数,yi是第i种算法输出值,ai是第i种算法的系数,i是正整数。建立了基于多种生理信息的抑郁评估数学模型后,利用抑郁评估数学模型的输出结果评价抑郁等级,所述抑郁等级分为五级:正常、一般、轻度抑郁、中度抑郁和重度抑郁。
相对于现有技术,本发明具备以下的优点及有益效果:
1、抑郁评估数学模型的建立具有研究基础,心电信号、脉搏波信号、脑电信号、皮电信号、胃电信号、肌电信号、眼电信号、多导睡眠信号和温度信号的参数与抑郁相关,因此利用基于生理信息的抑郁评估数学模型的输出结果评估抑郁等级具有可行性;
2、利用抑郁评估数据模型的评估方式通过生理参数客观量化抑郁等级,能够改善传统量表评估抑郁的方式,避免量表评估的主观性,符合临床需求并具有临床实用性;3、本发明结合心电、脉搏波、脑电、皮电、胃电、肌电、眼电、多导睡眠和温度的生理参数对抑郁症进行评估,丰富了神经科学领域与心理学领域交叉研究的方法;
4、本发明对心电信号以及脉搏波信号、脑电信号、皮电信号、胃电信号、肌电信号、眼电信号、多导睡眠信号和温度信号中的一种信号或一种以上的信号的结合进行信号处理、参数计算、建立数学模型,可选择多种信号组合进行评估,具有灵活性和新颖性;
5、本发明提出对信号参数归一化处理的方法,将参数与正常样本中的均值和标准差进行比较,消除参数在数值大小和偏差方面的差异,使参数集特征选择更加科学有效;
6、本发明提出多种特征选择和机器学习的算法组合,根据信号类型的不同,数学模型的建立方式更加灵活;
附图说明
图1为基于生理信息的抑郁症评估系统原理图。
图2为基于生理信息的抑郁症评估系统结构图。
具体实施方式
下面结合实施例及附图对本发明作进一步详细的描述,但本发明的实施方式不限于此。
实施例
如图1所示,一种基于生理信息的抑郁症评估系统,包括:信息采集模块、信号处理模块、参数计算模块、特征选择模块、机器学习模块、输出结果模块;信息采集模块采集的信号通过USB串口有线传输或者蓝牙无线传输的方式传输到信号处理模块中。信号处理模块输出经过处理的信号到参数计算模块。参数计算模块输出信号参数到特征选择模块。特征选择模块输出特征参数集到机器学习模块。机器学习模块输出抑郁等级到输出结果模块。
所述基于生理信息的抑郁症评估系统的结构如图2所示,所述的信息采集模块,用于采集心电信号并采集脉搏波信号、脑电信号、皮电信号、胃电信号、肌电信号、眼电信号、多导睡眠信号和温度信号中的一种信号或一种以上的信号。所述的信号处理模块,用于处理生理信息,包括去基线处理、滤波去噪处理、提取心搏间期处理、时频变换处理以及谱分析和谱估计处理等。所述的参数计算模块,用于计算经过处理的信号的信号参数,包括心率变异性的时域参数、频域参数和时域几何参数,以及根据采集的生理信息选择性地计算脉搏波信号、脑电信号、皮电信号、胃电信号、肌电信号、眼电信号、多导睡眠信号、温度信号中一种或一种以上信号的时域参数、频域参数、直方图参数、分布图参数。所述的特征选择模块,用于在全部信号参数中获取与抑郁等级相关的特征参数集。所述的机器学习模块,用于训练抑郁等级量化的分类器,利用特征参数集建立抑郁评估数学模型,量化抑郁等级。所述的输出结果模块,用于显示抑郁评估数学模型输出的抑郁等级。
该系统的基于多种生理信息的抑郁症评估方法具体实施步骤如下:
步骤1:获取生理信息,生理信息包括心电,以及脉搏波、脑电、皮电、胃电、肌电、眼电、多导睡眠、温度中一种或一种以上生理信息。其中:
心电信号采集可选择测量五分钟静息状态下的心电信号,心电采集采样率可以选择500Hz或者500Hz以上;
脉搏波采集可选择性利用红外光透射尖部位输出反应血管末稍血容积变化的脉搏传感器采集之间脉搏信号,或者利用震感式测量法采集腕部脉搏信号, 脉搏波采集采样率可以选择500Hz或者500Hz以上;
脑电采集可选择采用10-20系统点激发采集大脑皮层的自发脑电活动;
皮电采集采用皮肤交感反应测试,单脉冲经皮电刺激腕部正中神经,测试皮肤交感反应起始潜伏期和波幅,以及测试右手大鱼肌和前臂掌侧的皮肤电阻值;
胃电采集采用置于上腹部的体表电极测量胃肌电活动;
肌电采集采用生物反馈仪刺激,连接前额的肌电电极测量肌电的信号;
眼电采集采用闭眼眼球活动(CEM)测量;
多导睡眠采用同时采集眼电、下颌肌电和脑电的方式测量睡眠时间及其参数;
温度采集可采用红外测温原理测量体内温度的方式。信号采集属于常规信号采集。
步骤2:对步骤1获取的生理信息进行信号处理,计算信号参数;具体的参数列表如下表表1所示,表1为电信号及其参数描述列表:
Figure PCTCN2015093158-appb-000003
表1
其中,心电信号处理和参数计算通过心电信号去基线处理、滤波去噪处理、提取RR间期处理、插值处理、傅里叶变换处理以及谱分析和谱估计处理计算RR间期、Mean、SDNN、RMSSD、PNN50、SDSD、VLF、LF、HF、TP、LF/HF、SD1、SD2、a1和a2;
脉搏波信号处理和参数计算通过脉搏波信号去基线处理、滤波去噪处理、提取PP间期处理、插值处理、傅里叶变换处理以及谱分析和谱估计处理;
脑电信号处理和参数计算通过脑电信号去基线处理、阈值去噪处理、小波分解处理以及谱分析和谱估计处理计算δ波幅值、δ波功率、δ波均值、δ波方差、δ波偏歪度、δ波峭度、θ波幅值、θ波功率、θ波均值、θ波方差、θ波偏歪度、θ波峭度、α波幅值、α波功率、α波均值、α波方差、α波偏歪度、α波峭度、β波幅值、β波功率、β波均值、β波方差、β波偏歪度、β波峭度和小波熵;
皮电信号处理和参数计算通过皮电信号去基线处理和小波滤波计算皮肤交感反应潜伏期、皮肤交感反应波幅和皮肤电阻值;
胃电信号处理和参数计算通过胃电信号去基线处理、Hilbert-Huang变换处理、小波分析处理、多分辨率分析处理和独立成分分析处理计算正常胃电节律、慢波、胃动过缓和胃动过速成分;
肌电信号处理和参数计算通过肌电信号去基线处理和小波包自适应阈值去噪处理计算基础值、最小值、最高值、肌电下降能力和肌电曲线;
眼电信号处理和参数计算通过眼电信号去基线处理、加权中值滤波处理和小波变换处理计算R波成分、r波成分、S波成分和s波成分;
多导睡眠信号处理和参数计算通过睡眠脑电信号去基线处理、阈值去噪处理、小波分解处理以及谱分析和谱估计处理,睡眠眼电信号去基线处理、加权中值滤波处理和小波变换处理,睡眠肌电信号去基线处理、小波包自适应阈值去噪处理和睡眠分期处理计算睡眠潜伏期、睡眠总时间、觉醒指数、S1、S2、S3、S4、快速眼动百分比、快速眼动睡眠周期数、快速眼动睡眠潜伏期、快速眼动睡眠强度、快速眼动睡眠密度和快速眼动睡眠时间;
温度信号处理和参数计算通过温度信号去基线处理、阈值滤波处理和建立温度值与图像灰度值的关系式计算人体体内温度分布。
步骤3:利用步骤2计算得到的信号参数进行归一化处理,对经过归一化处理后的信号参数组成的参数集进行特征选择,得到特征参数集,所述的归一化处理方法:
Figure PCTCN2015093158-appb-000004
其中,X是指参数集的信号参数,Xi表示第i个进行归一化的信号参数值,Xin表示第i个归一化后的值,Ximean表示第i个参数的正常均值,Xistd表示第i个参数的正常标准差,i是正整数。所述的特征选择分为特征搜索和评价准则两部分,其中搜索算法使用以下算法中一种或一种以上的组合:完全搜索(Complete Search)、顺序搜索(Sequential Search)、随机搜索算法(Random Search)、遗传算法(Genetic Algorithm)、模拟退火搜索算法(Simulated Annealing)、可回溯的贪婪搜索扩张算法,评价准则可选择性地使用Wapper模型或CfsSubsetEval属性评估方法。其中在评估过程中获取心电和脉搏波信号,特征选择采用结合完全搜索算法与Wapper模型的方式;评估过程中,获取心电、皮电和多导睡眠信号,特征选择采用结合随机搜索算法与CfsSubsetEval属性评估方法的方式。根据采集信号种类不同,选择合适、准确度高的算法组合。
步骤4:根据步骤3得到的特征参数集进行机器学习,使用特征参数集在机器学习的过程中建立抑郁评估数学模型。其中机器学习的算法可选择性地使用以下算法中一种或一种以上组合:贝叶斯分类器(Bayes)、决策树算法(Decision Tree)、AdaBoost算法、k-近邻法(k-Nearest Neighbor )、支持向量机(SVM)。抑郁评估数学模型的表达式为:
Figure PCTCN2015093158-appb-000005
其中,Y是抑郁评估数学模型输出值,n是选择使用的机器学习算法数,yi是第i种算法输出值,ai是第i种算法的系数,i是正整数。所述抑郁评估数学模型建立了基于多种生理信息的抑郁评估数学模型后,利用抑郁评估数学模型的输出结果评价抑郁等级,把所述抑郁等级分为五级:正常、一般、轻度抑郁、中度抑郁和重度抑郁。
上述实施例为本发明较佳的实施方式,但本发明的实施方式并不受上述实施例的限制,其他的任何未背离本发明的精神实质与原理下所作的改变、修饰、替代、组合、简化,均应为等效的置换方式,都包含在本发明的保护范围之内。

Claims (9)

  1. 一种基于生理信息的抑郁症评估系统,其特征在于,包括:依次连接的信息采集模块、信号处理模块、参数计算模块、特征选择模块、机器学习模块和输出结果模块;
    信息采集模块,用于采集心电信号并采集脉搏波信号、脑电信号、皮电信号、胃电信号、肌电信号、眼电信号、多导睡眠信号和温度信号中的一种信号或一种以上的信号;信息采集模块采集的信号通过USB串口有线传输或蓝牙无线传输的方式传输到信号处理模块中;
    信号处理模块,用于处理生理信息,所述生理信息的处理包括去基线处理、滤波去噪处理、提取心搏间期处理、时频变换处理以及谱分析和谱估计处理,信号处理模块输出经过处理的信号到参数计算模块;
    参数计算模块,用于计算经过处理的信号的信号参数,所述信号参数包括心率变异性的时域参数、频域参数、时域几何参数以及根据采集的生理信息计算脉搏波信号、脑电信号、皮电信号、胃电信号、肌电信号、眼电信号、多导睡眠信号或温度信号中的一种或一种以上信号的时域参数、频域参数、直方图参数和分布图参数,参数计算模块输出信号参数到特征选择模块;
    特征选择模块,用于在全部信号参数中获取与抑郁等级相关的特征参数集,特征选择模块输出特征参数集到机器学习模块;
    机器学习模块,用于训练抑郁等级量化的分类器,利用特征参数集建立抑郁评估数学模型,量化抑郁等级,机器学习模块输出抑郁等级到输出结果模块;
    输出结果模块,用于显示抑郁评估数学模型输出的抑郁等级。
  2. 根据权利要求1所述的基于生理信息的抑郁症评估系统,其特征在于,所述的信息采集模块用于采集心电信号,所述的信息采集模块还用于采集心电信号采集脉搏波信号、脑电信号、皮电信号、胃电信号、肌电信号、眼电信号、多导睡眠信号和温度信号中的一种信号或一种以上的信号,所述的采集心电信号的采集方法采用三导联心电采集方法,在所述三导联心电采集方法中,采集到的心电信号经过放大、滤波和模数转换后,再通过数据传输将心电信号传输到电脑中,所述的数据传输采用USB串口有线传输或蓝牙无线传输。
  3. 根据权利要求1所述的基于生理信息的抑郁症评估系统,其特征在于,所述的信号处理模块包括:心电信号处理单元、脉搏波信号处理单元、脑电信 号处理单元、皮电信号处理单元、胃电信号处理单元、肌电信号处理单元、眼电信号处理单元、多导睡眠信号处理单元和温度信号处理单元;
    所述心电信号处理单元,用于去基线处理、滤波去噪处理、提取RR间期处理、插值处理、傅里叶变换处理以及谱分析和谱估计处理;
    所述脉搏波信号处理单元,用于去基线处理、滤波去噪处理、提取PP间期处理、插值处理、傅里叶变换处理以及谱分析和谱估计处理;
    所述脑电信号处理单元,用于去基线处理、阈值去噪处理、小波分解处理以及谱分析和谱估计处理;
    所述皮电信号处理单元,用于去基线处理和小波滤波处理;
    所述胃电信号处理单元,用于去基线处理、Hilbert-Huang变换处理、小波分析、多分辨率分析和独立成分分析;
    所述肌电信号处理单元,用于去基线处理和小波包自适应阈值去噪处理;
    所述眼电信号处理单元,用于去基线处理、加权中值滤波处理和小波变换处理;
    所述多导睡眠信号处理单元,用于处理睡眠脑电信号、睡眠眼电信号、睡眠肌电信号,对所述睡眠脑电信号进行去基线处理、阈值去噪处理、小波分解处理以及谱分析和谱估计处理,对所述睡眠眼电信号进行去基线处理、加权中值滤波处理和小波变换处理,对所述睡眠肌电信号进行去基线处理、小波包自适应阈值去噪处理和睡眠分期处理;
    所述温度信号处理单元,用于去基线处理、阈值滤波处理、建立温度值与图像灰度值的关系式和绘制人体热能分布图。
  4. 根据权利要求1所述的基于生理信息的抑郁症评估系统,其特征在于,所述的参数计算模块包括:心电参数计算单元、脉搏波参数计算单元、脑电参数计算单元、皮电参数计算单元、胃电参数计算单元、肌电参数计算单元、眼电参数计算单元、多导睡眠信号参数计算单元、温度参数计算单元;所述的心电参数计算单元包括:时域参数计算、频域参数计算和时域几何参数计算;
    所述心电参数计算单元,包括计算RR间期、时域参数、频域参数、和时域几何参数,所述时域参数包括:Mean、SDNN、RMSSD、PNN50和SDSD,所述频域参数包括:VLF、LF、HF、TP和LF/HF,所述时域几何参数包括:SD1、SD2、a1和a2;
    所述脉搏波参数计算单元,包括计算PP间期、时域参数、频域参数和时域 几何参数,所述时域参数包括Mean、SDNN、RMSSD、PNN50和SDSD,所述频域参数VLF、LF、HF、TP和LF/HF,所述时域几何参数包括SD1、SD2、a1和a2;
    所述脑电参数计算单元,用于计算δ波幅值、δ波功率、δ波均值、δ波方差、δ波偏歪度、δ波峭度、θ波幅值、θ波功率、θ波均值、θ波方差、θ波偏歪度、θ波峭度、α波幅值、α波功率、α波均值、α波方差、α波偏歪度、α波峭度、β波幅值、β波功率、β波均值、β波方差、β波偏歪度、β波峭度和小波熵;
    所述皮电参数计算单元,用于计算皮肤交感反应潜伏期、皮肤交感反应波幅和皮肤电阻值;
    所述胃电参数计算单元,用于计算正常胃电节律、慢波、胃动过缓成分和胃动过速成分;
    所述肌电参数计算单元,用于计算基础值、最小值、最高值、肌电下降能力和肌电曲线;
    所述眼电参数计算单元,用于计算R波成分、r波成分、S波成分和s波成分;
    所述多导睡眠信号参数计算单元,用于计算睡眠潜伏期、睡眠总时间、觉醒指数、S1、S2、S3、S4、快速眼动百分比、快速眼动睡眠周期数、快速眼动睡眠潜伏期、快速眼动睡眠强度、快速眼动睡眠密度和快速眼动睡眠时间;
    所述温度参数计算单元,用于计算人体体内温度分布和绘制人体热能图。
  5. 一种应用于权利要求1所述的基于生理信息的抑郁症评估系统的评估方法,其特征在于,包括以下步骤:
    步骤1:对心电信号进行信号处理并同时对脉搏波信号、脑电信号、皮电信号、胃电信号、肌电信号、眼电信号、多导睡眠信号和温度信号中的一种信号或一种以上的信号进行信号处理,再计算经过处理的信号的信号参数;
    步骤2:利用步骤1计算得到的信号参数进行归一化处理,对经过归一化处理后的信号参数组成的参数集进行特征选择,得到特征参数集;
    步骤3:利用步骤2得到的特征参数集进行机器学习,利用所述的特征参数集与抑郁等级的关系建立抑郁评估数学模型,利用所述的抑郁评估数学模型输出的抑郁等级评估结果,根据所述的抑郁等级的评估结果评估抑郁等级;
    所述的机器学习用于训练抑郁评估数学模型,使用特征参数集在机器学习的过程中建立抑郁评估数学模型,所述机器学习的算法使用以下算法中一种或一种以上的组合:贝叶斯分类器、决策树算法、AdaBoost算法、k-近邻法、支 持向量机,所述的抑郁评估数学模型的表达式为:
    Figure PCTCN2015093158-appb-100001
    其中,Y是抑郁评估数学模型输出值,n是选择使用的机器学习算法数,yi是第i种算法输出值,ai是第i种算法的系数,i是正整数。
  6. 根据权利要求5所述的评估方法,其特征在于,步骤2中,所述归一化处理方法为:
    Figure PCTCN2015093158-appb-100002
    其中,X是指参数集的信号参数,Xi表示第i个进行归一化的信号参数值,Xin表示第i个归一化后的值,Ximean表示第i个参数的正常均值,Xistd表示第i个参数的正常标准差,i是正整数。
  7. 根据权利要求5所述的评估方法,其特征在于,在步骤1中,所述信号处理包括心电信号处理,脉搏波信号处理,脑电信号处理,皮电信号处理,胃电信号处理,肌电信号处理,眼电信号处理,多导睡眠信号处理和温度信号处理,所述心电信号处理包括去基线处理、滤波去噪处理、提取RR间期、插值处理、傅里叶变换处理以及谱分析和谱估计处理,所述脉搏波信号处理包括去基线处理、滤波去噪处理、提取PP间期、插值处理、傅里叶变换处理以及谱分析和谱估计处理,所述脑电信号处理包括去基线处理、阈值去噪处理、小波分解处理以及谱分析和谱估计处理,所述皮电信号处理包括去基线处理以及小波滤波处理,所述胃电信号处理包括去基线处理、Hilbert-Huang变换处理、小波分析、多分辨率分析以及独立成分分析,所述肌电信号处理包括去基线处理和小波包自适应阈值去噪处理,所述眼电信号处理包括去基线处理、加权中值滤波处理和小波变换处理,所述多导睡眠信号处理包括处理睡眠脑电信号、睡眠肌电信号和睡眠眼电信号,对所述睡眠脑电信号进行去基线处理、阈值去噪处理、小波分解处理以及谱分析和谱估计处理,对所述睡眠眼电信号进行去基线处理、加权中值滤波处理和小波变换处理,对所述睡眠肌电信号进行去基线处理、小波包自适应阈值去噪处理和睡眠分期处理,所述温度信号处理包括去基线处理、阈值滤波处理和建立温度值与图像灰度值的关系式。
  8. 根据权利要求5所述的评估方法,其特征在于,在步骤1中,所述计算经过处理的信号的信号参数包括心电参数计算、脉搏波参数计算、脑电参数计 算、皮电参数计算、胃电参数计算、肌电参数计算、眼电参数计算、多导睡眠参数计算和温度参数计算,所述心电参数计算包括计算RR间期、时域参数、频域参数、和时域几何参数,所述时域参数包括Mean、SDNN、RMSSD、PNN50和SDSD,所述频域参数包括VLF、LF、HF、TP和LF/HF,所述时域几何参数包括SD1、SD2、a1和a2,所述脉搏波参数计算包括计算PP间期、时域参数、频域参数和时域几何参数,所述时域参数Mean、SDNN、RMSSD、PNN50、SDSD,所述频域参数包括VLF、LF、HF、TP和LF/HF,所述时域几何参数包括SD1、SD2、a1和a2,所述脑电参数计算包括计算δ波幅值、δ波功率、δ波均值、δ波方差、δ波偏歪度、δ波峭度、θ波幅值、θ波功率、θ波均值、θ波方差、θ波偏歪度、θ波峭度、α波幅值、α波功率、α波均值、α波方差、α波偏歪度、α波峭度、β波幅值、β波功率、β波均值、β波方差、β波偏歪度、β波峭度和小波熵,所述皮电参数计算包括计算皮肤交感反应潜伏期,皮肤交感反应波幅和皮肤电阻值,所述胃电参数计算包括计算正常胃电节律、慢波、胃动过缓和胃动过速成分,所述肌电参数计算包括计算基础值、最小值、最高值、肌电下降能力和肌电曲线,所述眼电参数计算包括计算R波、r波、S波和s波成分,所述多导睡眠信号参数计算包括计算睡眠潜伏期、睡眠总时间、觉醒指数、S1、S2、S3、S4、快速眼动百分比、快速眼动睡眠周期数、快速眼动睡眠潜伏期、快速眼动睡眠强度、快速眼动睡眠密度和快速眼动睡眠时间,所述温度参数计算包括计算体内温度分布。
  9. 根据权利要求5所述的评估方法,其特征在于,在步骤2中,所述特征选择根据参数计算模块输出的所有信号参数,训练数据集,每个样本用特征集表示,生成特征子集集合,根据评价准则搜索获取特征集中最好的特征子集,比较和评价当前的特征子集,当获取的特征子集是最好的特征子集,满足终止条件,输出与抑郁等级相关的特征参数集,所述搜索算法使用以下算法中一种或一种以上的组合:完全搜索算法、顺序搜索算法、随机搜索算法、遗传算法、模拟退火搜索算法和可回溯的贪婪搜索扩张算法;评价准则使用以下算法中一种或两种的组合:Wapper模型和CfsSubsetEval属性评估方法。
PCT/CN2015/093158 2015-07-30 2015-10-29 基于生理信息的抑郁症评估系统及其评估方法 WO2017016086A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US15/109,815 US20170238858A1 (en) 2015-07-30 2015-10-29 Depression assessment system and depression assessment method based on physiological information

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510468922.XA CN105147248B (zh) 2015-07-30 2015-07-30 基于生理信息的抑郁症评估系统及其评估方法
CN201510468922.X 2015-07-30

Publications (1)

Publication Number Publication Date
WO2017016086A1 true WO2017016086A1 (zh) 2017-02-02

Family

ID=54788561

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/093158 WO2017016086A1 (zh) 2015-07-30 2015-10-29 基于生理信息的抑郁症评估系统及其评估方法

Country Status (3)

Country Link
US (1) US20170238858A1 (zh)
CN (1) CN105147248B (zh)
WO (1) WO2017016086A1 (zh)

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018227239A1 (en) * 2017-06-12 2018-12-20 Medibio Limited Mental state indicator
CN109791564A (zh) * 2017-07-21 2019-05-21 深圳市汇顶科技股份有限公司 信号计算法中的参数的设定方法及装置
CN110236572A (zh) * 2019-05-07 2019-09-17 平安科技(深圳)有限公司 基于体温信息的抑郁症预测系统
CN111466910A (zh) * 2020-04-30 2020-07-31 电子科技大学 一种睡眠监测方法、系统、存储介质、计算机程序、装置
CN112568912A (zh) * 2019-09-12 2021-03-30 陈盛博 一种基于非侵入式脑电信号的抑郁症生物标记物辨识方法
CN113907768A (zh) * 2021-10-12 2022-01-11 浙江汉德瑞智能科技有限公司 一种基于matlab的脑电信号处理装置
WO2022095331A1 (zh) * 2020-11-09 2022-05-12 平安科技(深圳)有限公司 压力评估方法、装置、计算机设备和存储介质
CN115064246A (zh) * 2022-08-18 2022-09-16 山东第一医科大学附属省立医院(山东省立医院) 一种基于多模态信息融合的抑郁症评估系统及设备
CN115644872A (zh) * 2022-10-26 2023-01-31 广州建友信息科技有限公司 一种情绪识别方法、装置及介质
CN115778389A (zh) * 2022-12-02 2023-03-14 复旦大学 基于心电和皮肤电联合分析的分娩恐惧检测方法和系统
CN117289804A (zh) * 2023-11-23 2023-12-26 北京健康有益科技有限公司 虚拟数字人面部表情管理方法、装置、电子设备及介质

Families Citing this family (80)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105943065A (zh) * 2016-06-29 2016-09-21 北京工业大学 一种基于脑信息学系统化方法学的人体可穿戴生理-心理-行为数据采集与分析系统
CN109922726A (zh) * 2016-09-20 2019-06-21 夏普株式会社 状态取得计算机、状态取得方法以及信息处理系统
CN106333677A (zh) * 2016-09-21 2017-01-18 广州视源电子科技股份有限公司 睡眠状态分析中眨眼活动的检测方法和系统
CN106388778B (zh) * 2016-09-21 2019-06-11 广州视源电子科技股份有限公司 睡眠状态分析中的脑电信号预处理方法和系统
CN106551691B (zh) * 2016-12-02 2020-01-21 清华大学 一种心率变异性分析方法、装置及用途
CN106725535A (zh) * 2016-12-30 2017-05-31 中国科学院心理研究所 一种便携式惊吓反射仪及其操作方法
CN108320778A (zh) * 2017-01-16 2018-07-24 医渡云(北京)技术有限公司 病案icd编码方法及系统
CN106859617A (zh) * 2017-03-01 2017-06-20 浙江大学 一种穿戴式多生命体征参数采集设备及其参数提取方法
CN106618611A (zh) * 2017-03-06 2017-05-10 兰州大学 基于睡眠多通道生理信号的抑郁症辅助诊断方法和系统
CN107411734A (zh) * 2017-03-06 2017-12-01 华斌 一种根据人体生物电磁波获取用户特征的装置
JP6847721B2 (ja) * 2017-03-14 2021-03-24 オムロン株式会社 情報処理装置、情報処理方法及びそのプログラム
CN107007291A (zh) * 2017-04-05 2017-08-04 天津大学 基于多生理参数的紧张情绪强度识别系统及信息处理方法
CN107170443A (zh) * 2017-05-12 2017-09-15 北京理工大学 一种模型训练层AdaBoost算法的参数优化方法
CN109394203A (zh) * 2017-08-18 2019-03-01 广州市惠爱医院 精神障碍康复期情绪监测与干预方法
JP6927491B2 (ja) * 2017-09-12 2021-09-01 東洋紡株式会社 精神神経状態を判別する指標の作成方法および作成装置
JP6927492B2 (ja) * 2017-09-12 2021-09-01 東洋紡株式会社 睡眠障害を判別する指標の作成方法および作成装置
JP6865438B2 (ja) * 2017-09-12 2021-04-28 東洋紡株式会社 精神神経状態を判別する指標の作成方法および作成装置
CN107582037A (zh) * 2017-09-30 2018-01-16 深圳前海全民健康科技有限公司 基于脉搏波设计医疗产品的方法
CN107802273A (zh) * 2017-11-21 2018-03-16 重庆邮电大学 一种抑郁状态监测装置、系统及预测方法
CN107874750B (zh) * 2017-11-28 2020-01-10 华南理工大学 脉率变异性和睡眠质量融合的心理压力监测方法及装置
CN108492875A (zh) * 2018-02-07 2018-09-04 苏州中科先进技术研究院有限公司 一种用于康复治疗的系统、及其健康状态评估方法和装置
CN108577865B (zh) * 2018-03-14 2022-02-22 天使智心(北京)科技有限公司 一种心理状态确定方法及装置
CN108804246A (zh) * 2018-06-11 2018-11-13 上海理工大学 上肢康复机器人的可用性评价方法
CN109077714B (zh) * 2018-07-05 2021-03-23 广州视源电子科技股份有限公司 信号识别方法、装置、设备和存储介质
CN109199411B (zh) * 2018-09-28 2021-04-09 南京工程学院 基于模型融合的案件知情者识别方法
CN109363670A (zh) * 2018-11-13 2019-02-22 杭州电子科技大学 一种基于睡眠监测的抑郁症智能检测方法
CN109784023B (zh) * 2018-11-28 2022-02-25 西安电子科技大学 基于深度学习的稳态视觉诱发脑电身份识别方法及系统
CN109620259B (zh) * 2018-12-04 2020-10-27 北京大学 基于眼动技术与机器学习对孤独症儿童自动识别的系统
WO2020122227A1 (ja) * 2018-12-14 2020-06-18 学校法人慶應義塾 うつ状態を推定する装置、方法及びそのためのプログラム
CN109859570A (zh) * 2018-12-24 2019-06-07 中国电子科技集团公司电子科学研究院 一种大脑训练方法及系统
CN109620265A (zh) * 2018-12-26 2019-04-16 中国科学院深圳先进技术研究院 识别方法及相关装置
US20200205712A1 (en) * 2018-12-28 2020-07-02 X Development Llc Assessment of risk for major depressive disorder from human electroencephalography using machine learned model
CN111374647A (zh) * 2018-12-29 2020-07-07 中兴通讯股份有限公司 一种检测脉搏波的方法、装置和电子设备
CN109620266B (zh) * 2018-12-29 2021-12-21 中国科学院深圳先进技术研究院 个体焦虑水平的检测方法和系统
CN109685156B (zh) * 2018-12-30 2021-11-05 杭州灿八科技有限公司 一种用于识别情绪的分类器的获取方法
CN109875579A (zh) * 2019-02-28 2019-06-14 京东方科技集团股份有限公司 情绪健康管理系统和情绪健康管理方法
CN109938723A (zh) * 2019-03-08 2019-06-28 度特斯(大连)实业有限公司 一种人体疾病风险的判别方法及设备
CN110013250B (zh) * 2019-04-30 2021-08-17 中南大学湘雅二医院 一种抑郁症自杀行为的多模式特征信息融合预测方法
US11200814B2 (en) * 2019-06-03 2021-12-14 Kpn Innovations, Llc Methods and systems for self-fulfillment of a dietary request
US11205140B2 (en) * 2019-06-03 2021-12-21 Kpn Innovations Llc Methods and systems for self-fulfillment of an alimentary instruction set based on vibrant constitutional guidance
CN110599442B (zh) * 2019-07-01 2022-08-12 兰州大学 一种融合脑皮质厚度和边缘系统形态特征的抑郁识别系统
CN110292378B (zh) * 2019-07-02 2021-02-23 燕山大学 基于脑波闭环监测的抑郁症远程康复系统
CN110353704B (zh) * 2019-07-12 2022-02-01 东南大学 基于穿戴式心电监测的情绪评估方法与装置
CN110464367B (zh) * 2019-08-06 2021-11-23 合肥工业大学 基于多通道协同的心理异常检测方法和系统
KR102152957B1 (ko) 2019-10-18 2020-09-07 (의료)길의료재단 심박 변이도(hrv)에 기초한 기타 불안장애로부터 공황장애 감별 방법 및 그 장치
JP7285763B2 (ja) * 2019-11-18 2023-06-02 株式会社東海理化電機製作所 学習装置、学習方法、および測定装置
CN110946562B (zh) * 2019-11-25 2022-12-23 南京摩尼电子科技有限公司 基于Micro:bit微处理器的生理电信号测量分析方法及系统
US11464443B2 (en) * 2019-11-26 2022-10-11 The Chinese University Of Hong Kong Methods based on an analysis of drawing behavior changes for cognitive dysfunction screening
CN110916631B (zh) * 2019-12-13 2022-04-22 东南大学 基于可穿戴生理信号监测的学生课堂学习状态评测系统
CN111150411B (zh) * 2020-01-17 2022-11-11 哈尔滨工业大学 基于改进遗传算法的心理压力评测分级方法
CN111150410B (zh) * 2020-01-17 2022-11-11 哈尔滨工业大学 基于心电信号与肌电信号融合的心理压力评测方法
CN111248928A (zh) * 2020-01-20 2020-06-09 北京津发科技股份有限公司 压力识别方法及装置
CN111345800B (zh) * 2020-03-16 2022-11-01 华中师范大学 一种mooc环境下的学习注意力检测方法及系统
CN111588391A (zh) * 2020-05-29 2020-08-28 京东方科技集团股份有限公司 一种基于用户睡眠特征的精神状态确定方法及系统
CN111671423B (zh) * 2020-06-18 2022-02-18 四川大学 一种eeg信号的表示方法、分类方法、可视化方法及介质
TWI790479B (zh) * 2020-09-17 2023-01-21 宏碁股份有限公司 生理狀態評估方法與生理狀態評估裝置
CN112806994A (zh) * 2021-01-27 2021-05-18 首都师范大学 一种基于生理信号预测个体压力应对方式的系统和方法
CN112826451A (zh) * 2021-03-05 2021-05-25 中山大学 一种麻醉深度及睡眠深度的评估方法及装置
CN113057634A (zh) * 2021-03-29 2021-07-02 山东思正信息科技有限公司 一种心理测评与心电数据联合采集与处理方法及系统
CN113197585B (zh) * 2021-04-01 2022-02-18 燕山大学 一种神经肌肉信息交互模型构建及参数辨识优化方法
CN113633287A (zh) * 2021-07-08 2021-11-12 上海市精神卫生中心(上海市心理咨询培训中心) 一种基于语音分析的抑郁症识别方法、系统和设备
CN113397565A (zh) * 2021-07-22 2021-09-17 北京脑陆科技有限公司 基于脑电信号的抑郁识别方法、装置、终端及介质
CN113397563A (zh) * 2021-07-22 2021-09-17 北京脑陆科技有限公司 抑郁分类模型的训练方法、装置、终端及介质
CN113951905B (zh) * 2021-10-20 2023-10-31 天津大学 一种用于日常动态监测的多通道胃电采集系统
CN114010194A (zh) * 2021-11-03 2022-02-08 瑞尔明康(杭州)医疗科技有限公司 生物特征信息获取方法、装置及抑郁症评估装置
CN113974630A (zh) * 2021-11-26 2022-01-28 浙江昊梦科技有限公司 一种心理健康的检测方法及装置
CN115054248B (zh) * 2021-12-10 2023-10-20 荣耀终端有限公司 情绪监测方法和情绪监测装置
CN114305418B (zh) * 2021-12-16 2023-08-04 广东工业大学 一种用于抑郁状态智能评估的数据采集系统及方法
CN114081494B (zh) * 2022-01-21 2022-05-06 浙江大学 一种基于大脑外侧缰核信号的抑郁状态检测系统
CN114869298B (zh) * 2022-06-15 2024-07-02 浙大宁波理工学院 一种基于脑电信号的抑郁检测方法、系统及可存储介质
CN115349861A (zh) * 2022-08-23 2022-11-18 山东大学 一种基于单通道脑电信号的精神压力检测系统及方法
CN115363586A (zh) * 2022-09-08 2022-11-22 山东大学 一种基于脉搏波信号的心理压力等级评估系统及方法
CN115399773A (zh) * 2022-09-14 2022-11-29 山东大学 基于深度学习与脉搏信号的抑郁状态识别系统
CN115588484A (zh) * 2022-09-20 2023-01-10 北京中科心研科技有限公司 一种基于时间压力数学题任务的抑郁倾向识别系统
CN115568853A (zh) * 2022-09-26 2023-01-06 山东大学 一种基于皮电信号的心理压力状态评估方法及系统
CN115886818B (zh) * 2022-11-25 2024-02-09 四川大学华西医院 一种基于胃肠电信号的抑郁焦虑障碍预测系统及其构建方法
CN116189912A (zh) * 2023-04-25 2023-05-30 青岛市妇女儿童医院(青岛市妇幼保健院、青岛市残疾儿童医疗康复中心、青岛市新生儿疾病筛查中心) 一种具有学习功能的妇科患者生理信息反馈系统
CN116671881A (zh) * 2023-08-03 2023-09-01 北京九叁有方物联网科技有限公司 基于图神经网络的头戴式脑体作业能力评估设备及方法
CN117942076A (zh) * 2024-01-18 2024-04-30 好心情健康产业集团有限公司 基于单导脑电信号的心理状态识别方法及装置
CN117711626A (zh) * 2024-02-05 2024-03-15 江西中医药大学 一种基于多维度因素的抑郁情绪评测方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090082691A1 (en) * 2007-09-26 2009-03-26 Medtronic, Inc. Frequency selective monitoring of physiological signals
CN101642368A (zh) * 2008-08-04 2010-02-10 南京大学 自主神经功能信号的处理方法、装置和测试系统
CN103479349A (zh) * 2013-09-25 2014-01-01 深圳市理邦精密仪器股份有限公司 心电信号数据获取及处理方法和系统
CN104127194A (zh) * 2014-07-14 2014-11-05 华南理工大学 一种基于心率变异性分析方法的抑郁症的评估系统及方法
CN204274481U (zh) * 2014-07-14 2015-04-22 华南理工大学 一种抑郁症程度量化的评估系统
CN204931634U (zh) * 2015-07-30 2016-01-06 华南理工大学 基于生理信息的抑郁症评估系统

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080091090A1 (en) * 2006-10-12 2008-04-17 Kenneth Shane Guillory Self-contained surface physiological monitor with adhesive attachment
CN101917898A (zh) * 2007-10-31 2010-12-15 埃姆申塞公司 对来自观众的生理响应提供分散式收集和集中式处理的系统和方法
EP2057942B1 (en) * 2007-11-12 2012-05-16 Werner Bystricky Modeling the electrical activity of the heart by a single dipole, concurrently estimating subject and measurement related conditions
US20140057232A1 (en) * 2011-04-04 2014-02-27 Daniel Z. Wetmore Apparatus, system, and method for modulating consolidation of memory during sleep
US9189599B2 (en) * 2011-05-13 2015-11-17 Fujitsu Limited Calculating and monitoring a composite stress index
US20170215782A1 (en) * 2014-08-26 2017-08-03 Toyobo Co., Ltd. Method for determining a depression state and depression state determination device
US11076763B2 (en) * 2014-10-15 2021-08-03 Atlasense Biomed Ltd. Remote physiological monitor

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090082691A1 (en) * 2007-09-26 2009-03-26 Medtronic, Inc. Frequency selective monitoring of physiological signals
CN101642368A (zh) * 2008-08-04 2010-02-10 南京大学 自主神经功能信号的处理方法、装置和测试系统
CN103479349A (zh) * 2013-09-25 2014-01-01 深圳市理邦精密仪器股份有限公司 心电信号数据获取及处理方法和系统
CN104127194A (zh) * 2014-07-14 2014-11-05 华南理工大学 一种基于心率变异性分析方法的抑郁症的评估系统及方法
CN204274481U (zh) * 2014-07-14 2015-04-22 华南理工大学 一种抑郁症程度量化的评估系统
CN204931634U (zh) * 2015-07-30 2016-01-06 华南理工大学 基于生理信息的抑郁症评估系统

Cited By (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018227239A1 (en) * 2017-06-12 2018-12-20 Medibio Limited Mental state indicator
CN109791564B (zh) * 2017-07-21 2023-06-16 深圳市汇顶科技股份有限公司 信号计算法中的参数的设定方法及装置
CN109791564A (zh) * 2017-07-21 2019-05-21 深圳市汇顶科技股份有限公司 信号计算法中的参数的设定方法及装置
CN110236572A (zh) * 2019-05-07 2019-09-17 平安科技(深圳)有限公司 基于体温信息的抑郁症预测系统
CN112568912B (zh) * 2019-09-12 2024-05-14 江西盛梦科技有限公司 一种基于非侵入式脑电信号的抑郁症生物标记物辨识方法
CN112568912A (zh) * 2019-09-12 2021-03-30 陈盛博 一种基于非侵入式脑电信号的抑郁症生物标记物辨识方法
CN111466910B (zh) * 2020-04-30 2023-11-21 电子科技大学 一种睡眠监测方法、系统、存储介质、计算机程序、装置
CN111466910A (zh) * 2020-04-30 2020-07-31 电子科技大学 一种睡眠监测方法、系统、存储介质、计算机程序、装置
WO2022095331A1 (zh) * 2020-11-09 2022-05-12 平安科技(深圳)有限公司 压力评估方法、装置、计算机设备和存储介质
CN113907768A (zh) * 2021-10-12 2022-01-11 浙江汉德瑞智能科技有限公司 一种基于matlab的脑电信号处理装置
CN115064246A (zh) * 2022-08-18 2022-09-16 山东第一医科大学附属省立医院(山东省立医院) 一种基于多模态信息融合的抑郁症评估系统及设备
CN115064246B (zh) * 2022-08-18 2022-12-20 山东第一医科大学附属省立医院(山东省立医院) 一种基于多模态信息融合的抑郁症评估系统及设备
CN115644872A (zh) * 2022-10-26 2023-01-31 广州建友信息科技有限公司 一种情绪识别方法、装置及介质
CN115778389A (zh) * 2022-12-02 2023-03-14 复旦大学 基于心电和皮肤电联合分析的分娩恐惧检测方法和系统
CN115778389B (zh) * 2022-12-02 2024-05-28 复旦大学 基于心电和皮肤电联合分析的分娩恐惧检测方法和系统
CN117289804A (zh) * 2023-11-23 2023-12-26 北京健康有益科技有限公司 虚拟数字人面部表情管理方法、装置、电子设备及介质
CN117289804B (zh) * 2023-11-23 2024-02-13 北京健康有益科技有限公司 虚拟数字人面部表情管理方法、装置、电子设备及介质

Also Published As

Publication number Publication date
CN105147248B (zh) 2019-02-05
CN105147248A (zh) 2015-12-16
US20170238858A1 (en) 2017-08-24

Similar Documents

Publication Publication Date Title
WO2017016086A1 (zh) 基于生理信息的抑郁症评估系统及其评估方法
CN204931634U (zh) 基于生理信息的抑郁症评估系统
Montesinos et al. Multi-modal acute stress recognition using off-the-shelf wearable devices
Porta et al. Temporal asymmetries of short-term heart period variability are linked to autonomic regulation
Guo et al. Short-term analysis of heart rate variability for emotion recognition via a wearable ECG device
Ishaque et al. Physiological signal analysis and classification of stress from virtual reality video game
CN114010171B (zh) 一种基于心跳数据的分类器设置方法
Bong et al. Analysis of electrocardiogram (ECG) signals for human emotional stress classification
Cecchi et al. Physical stimuli and emotions: EDA features analysis from a wrist-worn measurement sensor
Sano et al. Comparison of sleep-wake classification using electroencephalogram and wrist-worn multi-modal sensor data
Wang et al. Emotionsense: An adaptive emotion recognition system based on wearable smart devices
Benchekroun et al. Comparison of Stress Detection through ECG and PPG signals using a Random Forest-based Algorithm
Wu et al. Automatic sleep-stage scoring based on photoplethysmographic signals
Wang et al. A novel rapid assessment of mental stress by using PPG signals based on deep learning
Chu et al. Physiological signals based quantitative evaluation method of the pain
Li et al. Detection of muscle fatigue by fusion of agonist and synergistic muscle semg signals
Pirbhulal et al. Analysis of efficient biometric index using heart rate variability for remote monitoring of obstructive sleep apnea
Kumar et al. INSOMNet: Automated insomnia detection using scalogram and deep neural networks with ECG signals
Klein et al. Sleep stages classification using vital signals recordings
CN115486819B (zh) 一种感知觉神经通路多级联检测量化的方法、系统和装置
Tara et al. Advances of cardiac state-inducing prototype and design of GUI to anatomize cardiac signal for ascertaining psychological working competence
Wu et al. Age-related differences in complexity during handgrip control using multiscale entropy
TWI516247B (zh) 居家式憂鬱傾向生理訊號之情緒分析方法
Farhan et al. Linear analysis of ECG data variability to assess the autonomic nervous system in two different body positions
CN116211308A (zh) 一种高强度运动下机体疲劳评估方法

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 15109815

Country of ref document: US

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15899438

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205 DATED 06.06.2018)

122 Ep: pct application non-entry in european phase

Ref document number: 15899438

Country of ref document: EP

Kind code of ref document: A1