CN112259237B - Depression evaluation system based on multi-emotion stimulus and multi-stage classification model - Google Patents
Depression evaluation system based on multi-emotion stimulus and multi-stage classification model Download PDFInfo
- Publication number
- CN112259237B CN112259237B CN202011092598.3A CN202011092598A CN112259237B CN 112259237 B CN112259237 B CN 112259237B CN 202011092598 A CN202011092598 A CN 202011092598A CN 112259237 B CN112259237 B CN 112259237B
- Authority
- CN
- China
- Prior art keywords
- subject
- depression
- information
- emotion
- stimulation
- Prior art date
- Legal status (The legal status 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 status listed.)
- Active
Links
- 238000011156 evaluation Methods 0.000 title claims abstract description 41
- 238000013145 classification model Methods 0.000 title claims abstract description 18
- 230000000638 stimulation Effects 0.000 claims abstract description 48
- 238000004140 cleaning Methods 0.000 claims abstract description 4
- 230000008451 emotion Effects 0.000 claims description 36
- 238000013210 evaluation model Methods 0.000 claims description 19
- 230000000994 depressogenic effect Effects 0.000 claims description 13
- 230000004424 eye movement Effects 0.000 claims description 9
- 230000008859 change Effects 0.000 claims description 7
- 230000007935 neutral effect Effects 0.000 claims description 7
- 210000004556 brain Anatomy 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 6
- 230000004936 stimulating effect Effects 0.000 claims description 6
- 238000012546 transfer Methods 0.000 claims description 5
- 230000036651 mood Effects 0.000 claims description 3
- 238000003745 diagnosis Methods 0.000 description 13
- 238000012360 testing method Methods 0.000 description 11
- 238000012549 training Methods 0.000 description 8
- 230000003001 depressive effect Effects 0.000 description 5
- 230000002996 emotional effect Effects 0.000 description 5
- 238000000034 method Methods 0.000 description 5
- 230000004044 response Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 3
- 230000002040 relaxant effect Effects 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000003925 brain function Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 230000003340 mental effect Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000000284 resting effect Effects 0.000 description 2
- 208000020401 Depressive disease Diseases 0.000 description 1
- 206010027951 Mood swings Diseases 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000007177 brain activity Effects 0.000 description 1
- 230000001914 calming effect Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000012631 diagnostic technique Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006397 emotional response Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000002650 habitual effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000004630 mental health Effects 0.000 description 1
- 230000006996 mental state Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001766 physiological effect Effects 0.000 description 1
- 230000006461 physiological response Effects 0.000 description 1
- 208000020016 psychiatric disease Diseases 0.000 description 1
- 238000001303 quality assessment method Methods 0.000 description 1
- 230000000087 stabilizing effect Effects 0.000 description 1
- 238000007920 subcutaneous administration Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/50—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for simulation or modelling of medical disorders
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/16—Devices for psychotechnics; Testing reaction times ; Devices for evaluating the psychological state
- A61B5/163—Devices for psychotechnics; Testing reaction times ; Devices for evaluating the psychological state by tracking eye movement, gaze, or pupil change
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/16—Devices for psychotechnics; Testing reaction times ; Devices for evaluating the psychological state
- A61B5/165—Evaluating the state of mind, e.g. depression, anxiety
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
Abstract
The embodiment of the invention discloses a depression evaluation system based on a multi-emotion stimulation and multi-level classification model, which comprises the following components: a stimulation module for performing exogenous and endogenous stimulation to the subject; the physiological signal acquisition module is used for acquiring target physiological information of the subject; the physiological signal analysis module is used for cleaning and extracting the characteristics of the target physiological information to obtain the target characteristic information of the subject; and the depression evaluation module is used for obtaining a depression evaluation result of the subject according to the target characteristic information. The invention can evaluate the subject more accurately and finely by collecting the physiological signals to be tested in all directions.
Description
Technical Field
The embodiment of the invention relates to the fields of mental health, physiological signal analysis and artificial intelligence aided diagnosis, in particular to a depression evaluation system based on multi-emotion stimulation and a multi-level classification model.
Background
Depression is a common psychological disorder. Depression is a major cause of disability worldwide and is a major factor in the worldwide burden of disease. Depression affects severely the quality of life and mental state of individuals, but in the current environment, the misdiagnosis rate and recurrence rate are high, and there is a lack of adequate fully trained relevant physicians for diagnosis and treatment.
The current diagnosis of depression is mainly based on clinical questionnaires, query observation of psychologists and subjective description of testers, and is easily influenced by prejudices and clinical experience of doctors, while some existing auxiliary diagnosis tools lack of judgment on the application range of the tools, so that the diagnosis is carried out on partial subjects easily misdiagnosed by the existing auxiliary diagnosis tools. There is therefore a need for a new objective clinical diagnostic technique.
Disclosure of Invention
Through a great deal of creative research, the inventor of the invention discovers that physiological responses of patients suffering from depression under brain activities, attention preferences and different emotion stimuli are greatly different from those of healthy individuals, and provides a theoretical basis for objectively evaluating depression tendency. The development progress of the current artificial intelligence and machine learning provides the possibility of realizing objective assessment of depression trend.
In the current field of auxiliary diagnosis and research of depression, physiological activities under single stimulus or resting state or multiple emotional stimuli with the same degree are generally collected, so that the stimulus may be too weak to generate enough response for partial patients with mild depression, or too strong to cause the healthy test to be induced to be close to the emotional response of the patients with depression. In addition, a single stimulus may increase the likelihood of false positives in the auxiliary diagnostic model due to the different depressive subtypes responding differently to the stimulus. At the same time, the subject is fully in a passive participation state in some of the disclosed schemes, which can lead to situations where the actual mental activities of the subject and the current stimuli may be isolated from each other. Especially in resting conditions, the state under test may be affected by a number of factors, such as experience prior to testing, which may interfere with the auxiliary diagnosis. There is therefore a need to organically combine multiple stimuli and increase the participation of the subject's initiative, thereby increasing the stability and reliability of the depression-aiding diagnostic system.
Some machine learning-based auxiliary diagnostic models of depression often only use existing data and algorithms to train an assessment model, but regardless of the limitations of the current test paradigm and the limitations of the number of tests, default to assessing all the tests and all possible subtypes of depression based on the current depression test paradigm. Meanwhile, given the differences between healthy and depressed populations, and the differences between the different depressed subtypes, the feature of being able to distinguish between depressed subtypes is not necessarily sufficiently effective in distinguishing between healthy and depressed populations, and vice versa. For the limitations of the existing auxiliary diagnosis model for depression, the evaluation of depression and the evaluation of depression subtype are required to be carried out separately, and the performance of the evaluation models per se is also required to be evaluated, namely whether the models have enough credibility to evaluate the currently tested depression state and depression subtype.
Based on the above knowledge, an object of an embodiment of the present invention is to provide a depression evaluation system based on multi-emotion stimulation and multi-level classification model, so as to solve the problem that the diagnosis of the existing depression is not objective.
In order to achieve the above purpose, the embodiment of the present invention mainly provides the following technical solutions:
the embodiment of the invention also provides a depression evaluation system based on the multi-emotion stimulation and the multi-level classification model, which comprises the following steps:
a stimulation module for performing exogenous and endogenous stimulation to the subject;
the physiological signal acquisition module is used for acquiring target physiological information of the subject;
the physiological signal analysis module is used for cleaning and extracting the characteristics of the target physiological information to obtain the target characteristic information of the subject;
and the depression evaluation module is used for obtaining a depression evaluation result of the subject according to the target characteristic information.
According to one embodiment of the invention, the stimulation module comprises:
a relaxation component for providing a first work of a mood that is flat and stable to the subject;
an endogenous stimulation component for endogenous stimulation of the subject;
and the exogenous stimulation component is used for stimulating the subject by at least one of neutral emotion stimulation, negative emotion stimulation and positive emotion stimulation in a picture and video mode or stimulating the subject by a combination of neutral emotion stimulation, negative emotion stimulation and positive emotion stimulation.
According to one embodiment of the invention, the physiological signal acquisition module comprises:
the electroencephalogram information acquisition component is used for acquiring and storing electroencephalogram signals of the subject;
a dermatologic information acquisition component for acquiring resistance change information at a second section of the index finger and middle finger of the non-dominant hand of the subject;
and the eye information acquisition component is used for acquiring eye information of the subject.
According to one embodiment of the invention, the ocular information collection component is further for performing a positional calibration prior to collecting ocular information of the subject.
According to one embodiment of the invention, the physiological signal analysis module is used for removing power frequency interference and noise interference from the electroencephalogram signals, then calculating a functional connection matrix of a target wave band under each stimulus and time-frequency information of the target wave band, and further obtaining difference values of corresponding conduction and connection corresponding characteristics of left and right brains according to the functional connection matrix and the time-frequency information of the target wave band, and the electroencephalogram signal variation characteristics under different degrees of different emotion stimuli.
According to one embodiment of the invention, the physiological signal analysis module is further configured to extract a detrend analysis dfa, entropy, mean value, standard deviation and parting dimension characteristics of the subcutaneous electrical signal under each emotion stimulus according to the resistance change information, and further compare the detrend analysis information, the entropy, the mean value and the standard deviation in each emotion dimension to obtain first differential characteristic information of each emotion stimulus.
According to one embodiment of the invention, the physiological signal analysis module is further configured to extract, according to eye movement information of the subject, attention point distribution information of the subject under each emotion stimulus, a tested longest gaze duration, and a gaze point transfer speed, and further obtain, according to the attention point distribution information of the subject under each emotion stimulus, the tested longest gaze duration, and the gaze point transfer speed, second differential feature information of the subject under each emotion stimulus.
According to one embodiment of the invention, the depression assessment module comprises:
the depression evaluation component is used for evaluating the target characteristic information of the subject input by the physiological signal analysis module according to a pre-trained deterministic evaluation model and a depression evaluation model to obtain a first unit result of depression evaluation of the subject;
the depression subtype evaluation component is used for evaluating the target characteristic information of the subject input by the physiological signal analysis module according to the pre-trained deterministic evaluation model and subtype analysis model to obtain a depression evaluation second unit result of the subject.
According to one embodiment of the invention, the usage of the depression assessment model is characterized by a significant difference between healthy and depressed populations.
According to one embodiment of the invention, the subtype analysis model is characterized by the use of a significant difference between the depressed subtypes.
The technical scheme provided by the embodiment of the invention has at least the following advantages:
the depression evaluation system based on the multi-emotion stimulation and the multi-stage classification model provided by the embodiment of the invention receives stimulation through exogenous and endogenous different-degree different-type emotion stimulation from shallow to deep, is tested to be subjected to three processes of passive participation, semi-active participation and active participation, and induces psychological-physiological responses of different degrees and types. The physiological signals to be tested are collected in an omnibearing way, so that the tested person can be evaluated more accurately and finely. The invention provides a multistage classification model architecture, and a set of pipelined auxiliary diagnosis model for depression is constructed based on the architecture. In this architecture, each level evaluates an aspect of the test, and each level itself evaluates whether the level has sufficient confidence; the low coupling between the levels, the next level only depends on the evaluation result of the previous level, and convenience is provided for updating a certain level in the future.
Drawings
Fig. 1 is a block diagram of a depression assessment system based on a multi-emotion stimulus and multi-level classification model in accordance with an embodiment of the present invention.
Detailed Description
Further advantages and effects of the present invention will become apparent to those skilled in the art from the disclosure of the present invention, which is described by the following specific examples.
In the following description, for purposes of explanation and not limitation, specific details are set forth such as the particular system architecture, interfaces, techniques, etc., in order to provide a thorough understanding of the present invention. It will be apparent, however, to one skilled in the art that the present invention may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, circuits, and methods are omitted so as not to obscure the description of the present invention with unnecessary detail.
In the description of the present invention, it is to be understood that the terms "first" and "second" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In the description of the present invention, it should be noted that, unless explicitly stated and limited otherwise, the terms "connected" and "connected" are to be construed broadly, and may be connected directly or indirectly through intermediaries, for example. The specific meaning of the above terms in the present invention will be understood in specific cases by those of ordinary skill in the art.
Fig. 1 is a block diagram of a depression assessment system based on a multi-emotion stimulus and multi-level classification model in accordance with an embodiment of the present invention. As shown in fig. 1, the depression assessment system based on the multi-emotion stimulation and multi-stage classification model according to the embodiment of the present invention includes a stimulation module 100, a physiological signal acquisition module 200, a physiological signal analysis module 300, and a depression assessment module 400.
The stimulation module 100 is used to collect target physiological information of a subject.
In one embodiment of the invention, the stimulation module 100 includes a relaxing component, an endogenous stimulation component, and an exogenous stimulation component.
In particular, the relaxation component provides relaxed video and background sounds for calming and stabilizing the subject's mood so that the subject can better enter a test state and at the end of the test relieve mood swings due to different emotional stimuli.
Exogenous stimulus components include pictorial and video stimuli, including stimuli in the neutral, negative and positive three emotional dimensions.
The subject firstly receives the affective stimulation of the pictures, is in light stimulation and passive participation, displays four pictures in a mode of 2 multiplied by 2 each time, and is three affective pictures (two negative pictures) or only neutral pictures, and when the three affective pictures are simultaneously present, the position of each affective picture is randomly placed so as to eliminate the influence of habitual watching positions. And pausing for a plurality of seconds between any two picture presentations for flattening the influence of the last picture stimulus. The pictures are randomly selected from the existing picture library, and after repeating 20 rounds, video stimulation is performed.
Compared with the picture stimulus, the video stimulus is more vivid, has substitution sense and has deeper stimulus degree. In the video stimulation stage, besides passive watching of the subject, the subject needs to actively participate, and after watching negative, positive or neutral videos each time, the video content needs to be recalled, so that psychological response of the subject is further enhanced.
The endogenous stimulus component can interact with the subject by way of speech in order to require the subject to recall past experiences, e.g., the endogenous stimulus component "please close the eyes" by speech, recall what you are most hard/happy in childhood. Because of the crafted experience, the subject will experience greater mental fluctuation than the exogenous stimulus, requiring recall to be both a sad experience and a happy experience.
In one embodiment of the invention, the depression assessment system based on the multi-emotional stimulus and multi-level classification model further comprises an adjustment module for adjusting the seat height of the subject and the height of the display device. The subject is positioned in front of the stimulation display device, the sitting posture is adjusted, the subject is positioned in a comfortable posture as much as possible, and meanwhile, the height of the display device and the height of the seat are adjusted, so that the subject and the stimulation display device are generally positioned in the same relative position. The subject receives the stimulus delivered by the relaxing component, the exogenous stimulating component, the endogenous stimulating component, and the relaxing component in the stimulating module in sequence.
The physiological signal acquisition module 200 is used for acquiring target physiological information of a subject.
In one embodiment of the invention, the physiological signal acquisition module 200 includes a brain electrical information acquisition component, a skin electrical information acquisition component, and an eye electrical information acquisition component.
The electroencephalogram information acquisition component is used for acquiring and storing electroencephalogram signals of a subject. In this embodiment, the electroencephalogram information acquisition component includes an electroencephalogram cap, the electroencephalogram cap is worn by a subject, the electrodes are wet electrodes, and then the electroencephalogram information acquisition component is connected to an electroencephalogram information acquisition device to acquire and store electroencephalogram information.
The dermatologic information acquisition component is used for acquiring resistance change information at the second sections of the index finger and the middle finger of the non-handy hand of the subject.
The oculogram information acquisition component is used for acquiring the oculogram information of the subject. In this embodiment, the eye electrical information collection assembly includes an eye movement collector. The eye movement collector is positioned in front of the stimulation display device and is used for collecting eye movement information. Before eye movement is collected, calibration is needed to ensure that the sight line position can be accurately collected.
It should be noted that all devices in the physiological signal acquisition module 200 that acquire signals need to synchronize time with the stimulation module 100 to ensure that physiological signals and different stimuli can be associated during subsequent analysis.
The physiological signal analysis module 300 is used for cleaning and extracting features of the target physiological information to obtain target feature information of the subject. Wherein the target characteristic information includes brain electrical signals, skin electrical signals, and eye movement information of the subject.
Specifically, for the electroencephalogram signal, the physiological signal analysis module 300 performs notch processing first to remove power frequency interference (50 Hz in China), and then removes noise interference such as electro-oculogram noise according to the electro-oculogram reference electrode. Performing different conductive connection electroencephalogram quality assessment, and performing interpolation processing on the poor electroencephalogram signals; and calculating the functional connection matrix of delta, theta, alpha, beta, gamma and Hgamma wave bands under each stimulus and the time-frequency information of each wave band. Because the brain function connection of the depressed patients is different, the brain function is sideways and the response to different emotion stimuli is different from that of normal people, the small world attribute, the rich-club attribute and the like of the function connection matrix of each stimulus stage are extracted, the power characteristics of each conduction electroencephalogram under different stimuli are extracted, the standard deviation, the average value, the entropy, the dfa (detritus analysis) and the fractal dimension characteristics of the electroencephalogram under different stimulus conditions and the power variation of the electroencephalogram with time are calculated, and the difference value of the corresponding characteristics of the corresponding conduction of the left brain and the right brain and the variation characteristics of the electroencephalogram under different emotion stimuli with different degrees are calculated.
Under the condition that the emotion state is changed, the skin conductivity of an individual is also changed, and the influence of a depressed patient on emotion stimulus is abnormal, and meanwhile, under the condition of different degrees of emotion stimulus, the change degree is different from that of a normal person, so that the emotion change of the subject can be analyzed according to skin electric signals, and whether the subject has depression tendency or not can be judged. For the skin electric signal, the physiological signal analysis module 300 extracts the characteristics of dfa, entropy, mean value, standard deviation, parting dimension and the like of the skin electric under different degrees of different emotion stimuli, and then compares the characteristics in each emotion dimension to obtain first differential characteristic information of different degrees of different emotion stimuli.
For eye movement information, the physiological signal analysis module 300 analyzes subject focus scene information, and based on existing research, it is considered that a depressed patient will focus more on negative scene characteristics, so that under various emotional stimulus conditions of different degrees, main focus distribution information to be tested, longest fixation time length to be tested, fixation point transfer speed and the like are extracted, and on the basis, second differential characteristic information under various emotional stimulus of different degrees is obtained.
The physiological signal analysis module 300 screens the characteristics obtained by the analysis of the electroencephalogram signals, the skin electric signals and the eye movement information, and the characteristics with significant differences between healthy people and depressive patients and the characteristics with significant differences between depressive subtypes are transmitted to the depression evaluation module 400 for analysis.
The depression evaluation module 400 is configured to obtain a depression evaluation result of the subject according to the target feature information.
Specifically, the depression assessment module 400 is a hierarchical classification model architecture, each level depends only on the data and features input by the physiological signal analysis module 300, and whether or not to execute the level depends only on the results of the next previous level, i.e., if the previous level executes and returns results requiring the next level to execute, the lower level module executes. Taking a secondary depression assessment as an example, the first level determines whether the subject is depressed and the second level assesses the subtype of depression being tested. Both depression and depression subtypes are completed at the discretion of multiple physicians during the model building process.
In one embodiment of the invention, depression assessment module 400 includes a depression assessment component and a depression subtype assessment component.
The depression evaluation component is used for evaluating the target characteristic information of the subject input by the physiological signal analysis module according to the pre-trained deterministic evaluation model and the depression evaluation model to obtain a first unit result of depression evaluation of the subject. The use of the depression assessment model is characterized by significant differences between healthy and depressed populations.
Specifically, using the lightgbm model, 80% of the training data was randomly extracted for model training for a total of 100 rounds. With the 100 models, a plurality of prediction results of each data are obtained, the average classification error is marked as difficult-to-classify data, and the rest is marked as sortable data.
And (3) randomly extracting 80% of data from the new label data, putting the data into the lightgbm for training, randomly carrying out 100 times to obtain 100 models, taking the 100 models as deterministic evaluation models, taking the average value of the predictive scores of the 100 models as the confidence level of the data which can be evaluated correctly by evaluating the depressive evaluation models, and carrying out depressive evaluation only when the confidence level is larger than a specified threshold value, otherwise, outputting the data without enough confidence to carry out auxiliary diagnosis of the depression.
And taking the correctly classified model as a training set to train a depression evaluation model. And randomly extracting 80% of data each time, putting the data into the lightgbm for training, repeating 100 times to obtain 100 models, combining the 100 models as a final depression evaluation model, and taking an average value of the model evaluation scores as a final depression tendency score, namely a first unit result of depression evaluation, when the depression evaluation is carried out.
The depression subtype evaluation component is used for evaluating the target characteristic information of the subject input by the physiological signal analysis module according to the pre-trained deterministic evaluation model and the subtype analysis model to obtain a depression evaluation second unit result of the subject.
Specifically, data marked as sortable are randomly extracted by 80%, data quantity balance among subtypes is kept in the extraction process, then the data are put into lightgbm for training, and the depressed subtypes are used as labels, and the process is carried out for 100 rounds; and comprehensively evaluating the prediction scores of each model, and marking the data with the scores lower than a threshold value as indistinguishable subtypes, and marking the data with the scores not lower than the threshold value as distinguishable subtypes.
The data are labeled with distinguishable subtypes and indistinguishable subtypes, and a deterministic assessment model is trained. 80% of data are randomly extracted each time, the data are put into the lightgbm for training, 100 times are repeated, and 100 obtained models jointly form a deterministic evaluation model. In the evaluation, the average value of the evaluation scores of the tested data in the hundred models is taken as the evaluation score of the deterministic model.
And randomly extracting 80% of data of analyzable subtypes, keeping data balance among the subtypes in the extraction process, putting the extracted data into a lightgbm model for training, taking each subtype as a label, repeating 100 times to obtain 100 models, and combining the 100 models to be used as a depression subtype evaluation model. In use, the average score of these 100 models is taken as the subtype score for the test, the subtype above the specified threshold being the subtype of depression tested.
It should be noted that, when the depression evaluation module 400 performs online auxiliary diagnosis, the tested data firstly enters a deterministic evaluation model in the depression evaluation component, if the score is lower than the threshold value, the current depression evaluation model is output without enough confidence to diagnose the tested depression trend, otherwise, the tested data enters a first-stage depression evaluation model to judge whether the tested data is depression, if the tested data is judged to be healthy, the tested data is directly output, otherwise, the tested data enters a second-stage depression subtype evaluation. In the second stage, firstly entering a deterministic assessment model, if the score is lower than a threshold value, outputting that the current model has insufficient confidence to assess the tested depression subtype, otherwise entering a subtype assessment model to obtain the tested depression subtype.
In addition, other configurations and functions of the depression assessment system based on the multi-emotion stimulation and multi-stage classification model according to the embodiments of the present invention are known to those skilled in the art, and are not described in detail for reducing redundancy.
The foregoing embodiments have been provided for the purpose of illustrating the general principles of the present invention in further detail, and are not to be construed as limiting the scope of the invention, but are merely intended to cover any modifications, equivalents, improvements, etc. based on the teachings of the invention.
Claims (4)
1. A depression assessment system based on a multi-emotional stimulus and multi-level classification model, comprising:
a stimulation module for performing exogenous and endogenous stimulation to the subject;
the physiological signal acquisition module is used for acquiring target physiological information of the subject;
the physiological signal analysis module is used for cleaning and extracting the characteristics of the target physiological information to obtain the target characteristic information of the subject;
the depression evaluation module is used for obtaining a depression evaluation result of the subject according to the target characteristic information;
the stimulation module includes:
a relaxation component for providing a first work of a mood that is flat and stable to the subject;
an endogenous stimulation component for endogenous stimulation of the subject;
the exogenous stimulation component is used for stimulating the subject in at least one mode of neutral emotion stimulation, negative emotion stimulation and positive emotion stimulation or a mode of combining neutral emotion stimulation, negative emotion stimulation and positive emotion stimulation in a mode of pictures and videos;
the physiological signal acquisition module comprises:
the electroencephalogram information acquisition component is used for acquiring and storing electroencephalogram signals of the subject;
a dermatologic information acquisition component for acquiring resistance change information at a second section of the index finger and middle finger of the non-dominant hand of the subject;
an eye information acquisition component for acquiring eye information of the subject;
the oculogram information acquisition component is further for performing a positional calibration prior to acquiring the oculogram information of the subject;
the physiological signal analysis module is used for removing power frequency interference and noise interference on the electroencephalogram signals, then calculating a functional connection matrix of a target wave band under each stimulus and time-frequency information of the target wave band, and further obtaining difference values of corresponding conduction corresponding characteristics of left and right brains and electroencephalogram signal variation characteristics under different-degree different-emotion stimuli according to the functional connection matrix and the time-frequency information of the target wave band;
the physiological signal analysis module is also used for extracting trending analysis dfa, entropy, mean value, standard deviation and parting dimension characteristics of the skin electric signal under each emotion stimulus according to the resistance change information, and further comparing the trending analysis information, the entropy, the mean value and the standard deviation in each emotion dimension to obtain first difference characteristic information of each emotion stimulus;
the physiological signal analysis module is further used for extracting the attention point distribution information, the tested longest gazing duration and the gazing point transfer speed of the subject under each emotion stimulus according to the eye movement information of the subject, and further obtaining second differential characteristic information of the subject under each emotion stimulus according to the attention point distribution information, the tested longest gazing duration and the gazing point transfer speed of the subject under each emotion stimulus.
2. The depression assessment system based on a multi-emotional stimulus and multi-level classification model of claim 1, wherein the depression assessment module comprises:
the depression evaluation component is used for evaluating the target characteristic information of the subject input by the physiological signal analysis module according to a pre-trained deterministic evaluation model and a depression evaluation model to obtain a first unit result of depression evaluation of the subject;
the depression subtype evaluation component is used for evaluating the target characteristic information of the subject input by the physiological signal analysis module according to the pre-trained deterministic evaluation model and subtype analysis model to obtain a depression evaluation second unit result of the subject.
3. The depression assessment system based on a multi-emotion stimulus and multi-stage classification model of claim 2, wherein the depression assessment model is characterized in use by significant differences between healthy and depressed populations.
4. A depression assessment system based on a multi-emotional stimulus and multi-level classification model according to claim 3, wherein the usage of the subtype analysis model is characterized by significant differences between the individual depression subtypes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011092598.3A CN112259237B (en) | 2020-10-13 | 2020-10-13 | Depression evaluation system based on multi-emotion stimulus and multi-stage classification model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011092598.3A CN112259237B (en) | 2020-10-13 | 2020-10-13 | Depression evaluation system based on multi-emotion stimulus and multi-stage classification model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112259237A CN112259237A (en) | 2021-01-22 |
| CN112259237B true CN112259237B (en) | 2024-03-26 |
Family
ID=74243062
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011092598.3A Active CN112259237B (en) | 2020-10-13 | 2020-10-13 | Depression evaluation system based on multi-emotion stimulus and multi-stage classification model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112259237B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113658697B (en) * | 2021-07-29 | 2023-01-31 | 北京科技大学 | Psychological assessment system based on video fixation difference |
| CN114246551A (en) * | 2021-11-01 | 2022-03-29 | 苏州市广济医院 | System for diagnosing schizophrenia and application thereof |
| CN115081486B (en) * | 2022-07-05 | 2023-07-04 | 华南师范大学 | System and method for positioning epileptic focus by using intracranial brain electrical network in early stage of epileptic seizure |
| CN115363585B (en) * | 2022-09-04 | 2023-05-23 | 北京中科心研科技有限公司 | Standardized group depression risk screening system and method based on habit removal and film watching tasks |
| CN115588484A (en) * | 2022-09-20 | 2023-01-10 | 北京中科心研科技有限公司 | Depression tendency recognition system based on time pressure mathematics subject task |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105592798A (en) * | 2013-03-15 | 2016-05-18 | 塞罗拉公司 | System and signatures for multi-modal physiological stimulation and assessment of brain health |
| CN108056774A (en) * | 2017-12-29 | 2018-05-22 | 中国人民解放军战略支援部队信息工程大学 | Experimental paradigm mood analysis implementation method and its device based on visual transmission material |
| CN109157231A (en) * | 2018-10-24 | 2019-01-08 | 阿呆科技（北京）有限公司 | Portable multi-channel Depression trend assessment system based on emotional distress task |
| CN109585021A (en) * | 2018-11-28 | 2019-04-05 | 华南理工大学 | State of mind assessment system and its working method based on line holographic projections technology |
| CN110522983A (en) * | 2018-05-23 | 2019-12-03 | 深圳先进技术研究院 | Brain stimulation system, method, equipment and storage medium based on artificial intelligence |
| CN110811648A (en) * | 2019-11-20 | 2020-02-21 | 阿呆科技（北京）有限公司 | Depression tendency evaluation system based on residual convolutional neural network |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10617351B2 (en) * | 2012-04-23 | 2020-04-14 | Sackett Solutions & Innovations Llc | Cognitive biometric systems to monitor emotions and stress |
| US20190385711A1 (en) * | 2018-06-19 | 2019-12-19 | Ellipsis Health, Inc. | Systems and methods for mental health assessment |
| US20200294670A1 (en) * | 2019-03-13 | 2020-09-17 | Monsoon Design Studios LLC | System and method for real-time estimation of emotional state of user |
-
2020
- 2020-10-13 CN CN202011092598.3A patent/CN112259237B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105592798A (en) * | 2013-03-15 | 2016-05-18 | 塞罗拉公司 | System and signatures for multi-modal physiological stimulation and assessment of brain health |
| CN108056774A (en) * | 2017-12-29 | 2018-05-22 | 中国人民解放军战略支援部队信息工程大学 | Experimental paradigm mood analysis implementation method and its device based on visual transmission material |
| CN110522983A (en) * | 2018-05-23 | 2019-12-03 | 深圳先进技术研究院 | Brain stimulation system, method, equipment and storage medium based on artificial intelligence |
| CN109157231A (en) * | 2018-10-24 | 2019-01-08 | 阿呆科技（北京）有限公司 | Portable multi-channel Depression trend assessment system based on emotional distress task |
| CN109585021A (en) * | 2018-11-28 | 2019-04-05 | 华南理工大学 | State of mind assessment system and its working method based on line holographic projections technology |
| CN110811648A (en) * | 2019-11-20 | 2020-02-21 | 阿呆科技（北京）有限公司 | Depression tendency evaluation system based on residual convolutional neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112259237A (en) | 2021-01-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112259237B (en) | Depression evaluation system based on multi-emotion stimulus and multi-stage classification model | |
| CN110801237B (en) | Cognitive ability evaluation system based on eye movement and electroencephalogram characteristics | |
| US20190391638A1 (en) | Methods and Systems for Obtaining, Aggregating, and Analyzing Vision Data to Assess a Person's Vision Performance | |
| KR101247748B1 (en) | Apparatus for analysing focus and nonfocus states and method therof | |
| US20150245800A1 (en) | Method for Detection Of An Abnormal Sleep Pattern In A Person | |
| Giannakakis et al. | A novel multi-kernel 1D convolutional neural network for stress recognition from ECG | |
| KR20150076167A (en) | Systems and methods for sensory and cognitive profiling | |
| Kim et al. | An analysis of smartphone overuse recognition in terms of emotions using brainwaves and deep learning | |
| CN113143208B (en) | Pain sensitivity assessment system and method based on multidimensional measurement | |
| CN112515676A (en) | Psychological examination method and subjective and objective combined examinee disguise testing method | |
| CN109199411A (en) | Case insider's recognition methods based on Model Fusion | |
| Bonotis et al. | Automated assessment of pain intensity based on EEG signal analysis | |
| CN112614583A (en) | Depression grade testing system | |
| CN114027842B (en) | Objective screening system, method and device for depression | |
| Desai et al. | Electrodermal activity (EDA) for treatment of neurological and psychiatric disorder patients: a review | |
| Nagai et al. | Comparing face processing strategies between typically-developed observers and observers with autism using sub-sampled-pixels presentation in response classification technique | |
| CN111743553B (en) | Emotion feature extraction method and system based on eye movement data | |
| Xia et al. | Dynamic viewing pattern analysis: towards large-scale screening of children with ASD in remote areas | |
| CN113397482A (en) | Human behavior analysis method and system | |
| CN112328072A (en) | Multi-mode character input system and method based on electroencephalogram and electrooculogram | |
| Parsons et al. | Neurogaming-based classification of player experience using consumer-grade electroencephalography | |
| Chen et al. | An effective entropy-assisted mind-wandering detection system with EEG signals based on MM-SART database | |
| Wu et al. | Stress detection using wearable devices based on transfer learning | |
| KR102494464B1 (en) | Smart glasses for monitoring brain nervous system diseases using eye recognition data | |
| CN113855022A (en) | Emotion evaluation method and device based on eye movement physiological signals |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |