A multi-modal depression disorder prediction system and method based on a viewing task

Abstract

The application discloses a kind of multi-modal depression disorder prediction system and method based on viewing task, system includes viewing task module, physiological signal acquisition equipment, feature extraction module, data processing module and prediction module, viewer wears wearable physiological signal acquisition equipment integrated with pulse wave sensor, skin resistance sensor;Viewing task module presents rest film segment, evokes acute psychological stress material segment and psychological intervention regulation material segment in turn for viewer;Synchronous acquisition physiological data signal when viewer watches each viewing task;The obtained physiological data signal is processed and extracted, and the feature difference I of the skin conductance level of the viewer in the viewing process and the feature difference I of high-frequency power spectral density are obtained;With judgment threshold value is compared and analyzed, the risk level of the depression disorder of viewer is predicted.The application combines film plot and physiological detection means, objectively analyzes the risk of existing depression disorder, and the detection accuracy is higher.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of depression disorder evaluation, and particularly relates to a multi-modal depression disorder prediction system and method based on a viewing task. BACKGROUND

[0002] With the continuous development of society, the pace of work and life of people is significantly accelerated, the competition pressure in daily life is increasing, and the incidence of depression disorder in the population is rising.

[0003] Depression disorder is also commonly known as melancholia or depression, which is a psychological disease with continuous and long-term low mood as the main clinical feature, and is the most important type of psychological disease of modern people. Long-term depression will seriously affect the social skills of patients, such as lack of self-confidence, actively avoiding crowds, and even suicidal tendencies and behaviors, and will also lead to disorders of the patient's body functions, such as sleep disorders and binge eating.

[0004] Clinicians make preliminary diagnosis according to common clinical symptoms, and can also diagnose depression of patients and assess the severity thereof through some professional standardized questionnaires. Unlike the detection methods of physiological diseases such as cancer and infectious diseases, structured questionnaires cannot accurately quantify the patient's condition through clear physiological indicator data or medical images. On the contrary, the professional level of clinicians and psychological consultants is highly dependent on observation and questionnaires, and the required diagnosis time is relatively long. In addition, due to the prejudice of society towards patients with mental illness, patients may deliberately conceal their condition, making the structured questionnaire detection inaccurate and wasting social resources. SUMMARY

[0005] In order to solve the above technical problems and realize objective prediction of patients with depression disorder, the present application provides a multi-modal depression disorder prediction system and method based on a viewing task, which combines viewing process and physiological parameter detection, and quickly predicts depression disorder of viewers according to changes in pulse wave signals and skin resistance signals.

[0006] The technical solutions adopted are as follows:

[0007] In one aspect, the present application provides a multi-modal depression disorder prediction system based on a viewing task, which comprises:

[0008] a viewing task module, which sequentially presents resting film materials, film materials for inducing acute psychological stress, and film materials for restoring physical and mental state in a set time sequence;

[0009] a physiological signal acquisition device, which is used to acquire pulse wave signals and skin resistance signals of viewers when performing the viewing task;

[0010] a feature extraction module configured to extract a high-frequency power spectral density feature from the collected pulse wave signal and a skin conductance level feature from the collected skin resistance signal;

[0011] a data processing module configured to calculate a feature difference I of the skin conductance level and a feature difference II of the high-frequency power spectral density during the viewing process of the viewer by using an inbuilt operation program;

[0012] a prediction module configured to set a judgment threshold I and a judgment threshold II, and configured to predict that the viewer has no risk of depression disorder and output when the feature difference I is higher than the judgment threshold I and the feature difference II is higher than the judgment threshold II, and configured to predict that the viewer has a risk of depression disorder and output when at least one of the feature difference I and the feature difference II is lower than the corresponding judgment threshold I and the judgment threshold II.

[0013] Further, the viewing task module comprises:

[0014] a resting viewing task unit configured to contain a calm and soothing resting film segment, and configured to enable the viewer to achieve physical and mental calmness during the viewing process;

[0015] a stress viewing task unit configured to contain a material segment for inducing acute psychological stress of the viewer, and configured to test the change of the sympathetic nerve activity of the viewer;

[0016] an intervention viewing task unit configured to contain a material segment for psychological intervention and adjustment of the viewer, and configured to enable the viewer to quickly recover the calmness and harmony of the physical and mental state, and to test the change of the parasympathetic nerve activity of the viewer.

[0017] Preferably, the physiological signal collection device collects the skin resistance signal I of the viewer when performing the resting viewing task, collects the pulse wave signal I and the skin resistance signal II of the viewer when performing the stress viewing task, and collects the pulse wave signal II of the viewer when performing the intervention viewing task.

[0018] The feature extraction module extracts the skin conductance level feature I in the skin resistance signal I, the skin conductance level feature II in the skin resistance signal II, the high-frequency power spectral density feature I in the pulse wave signal I, and the high-frequency power spectral density feature II in the pulse wave signal II in sequence.

[0019] Further preferably, the data processing module calculates the average value of the collected skin conductance level feature I, skin conductance level feature II, high frequency power spectral density feature I and high frequency power spectral density feature II respectively to obtain the resting skin resistance feature I, stress skin resistance feature II, stress heart rate variability high frequency feature I and intervention heart rate variability high frequency feature II; the feature difference value I is the difference between the stress skin resistance feature II and the resting skin resistance feature I, and the feature difference value II is the difference between the intervention heart rate variability high frequency feature II and the stress heart rate variability high frequency feature I.

[0020] Further preferably, the judgment threshold value I in the prediction module is the standard deviation calculated by the data processing module for the skin conductance level feature I, and the judgment threshold value II is the standard deviation calculated by the data processing module for the high frequency power spectral density feature II.

[0021] Further, the prediction module performs a graded prediction on the viewer who is predicted to have a risk of depression disorder: when the feature difference value I is higher than the judgment threshold value I, it is judged that the viewer has normal sympathetic nervous activity, and label A is output, otherwise, it is judged that the sympathetic nervous activity is abnormal, and label B is output.

[0022] When the feature difference value II is higher than the judgment threshold value II, it is judged that the viewer has normal parasympathetic nervous activity, and label C is output, otherwise, it is judged that the parasympathetic nervous activity is abnormal, and label D is output.

[0023] The prediction module also performs a graded prediction on the output label of depression disorder: when the viewer gets label A and label C at the same time, it is judged that the viewer has no risk of depression disorder, and is output.

[0024] When the viewer gets label B and label D at the same time, it is judged that the viewer has a risk of severe depression disorder, and is output.

[0025] When the viewer gets label A and label D, or label B and label C at the same time, it is judged that the viewer has a risk of mild depression disorder, and is output.

[0026] Preferably, the physiological signal acquisition device is a wearable pulse wave sensor and a skin resistance sensor, or a wearable acquisition device integrated with a pulse wave sensor and a skin resistance sensor.

[0027] In another aspect, the present application also provides a multi-modal depression disorder prediction method based on a viewing task. A viewer wears a wearable physiological signal acquisition device integrated with a pulse wave sensor and a skin resistance sensor. A viewing task module presents a rest film segment, a material segment for inducing acute psychological stress and a material segment for psychological intervention adjustment to the viewer in sequence. Physiological data signals of the viewer during the viewing of each viewing task are synchronously acquired. The physiological data signals of each viewing task are processed and features are extracted to obtain a feature difference I of a skin conductance level and a feature difference I of a high-frequency power spectral density of the viewer during the viewing. The depression disorder risk level of the viewer is predicted by comparing and analyzing the feature difference I with a set judgment threshold, and the depression disorder risk level of the viewer is output.

[0028] The technical scheme of the present application has the following advantages:

[0029] A. The viewing task module for presenting different film segments is arranged in the present application, so that the viewer receives physiological data acquisition during the viewing process. The sympathetic nerve and parasympathetic nerve signals reflecting the common features of depression disorder are collected, i.e. the skin resistance signals and pulse wave signals generated by the viewer when watching the set playback film are collected by the physiological signal acquisition device, the change values thereof during the entire viewing process are obtained, and the depression disorder risk is judged according to the change values. The depression disorder risk is objectively analyzed in combination with the film plot and the physiological detection means, and the detection accuracy is higher.

[0030] B. Compared with the traditional depression disorder detection and evaluation means in the form of time-consuming and lengthy questionnaire scale answering, the physiological evaluation means adopted in the present application has low task operation requirement, greatly reduces the difficulty and energy consumption of the test personnel during the evaluation, the data processing module has synchronous data analysis and operation capability, the evaluation result can be quickly obtained, the application scene of the depression disorder evaluation is expanded, and a simple and efficient evaluation mode is provided. BRIEF DESCRIPTION OF DRAWINGS

[0031] In order to more clearly illustrate the specific embodiments of the present application, the drawings needed in the specific embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0032] Figure 1 is a system structure diagram provided by the present application;

[0033] Figure 2 is a viewing task module composition diagram provided by the present application;

[0034] Figure 3 is a detection and discrimination analysis flowchart provided by the present application. DETAILED DESCRIPTION

[0035] The technical solutions of the present application will be described clearly and completely below in conjunction with the drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0036] As shown in Figure 1 and Figure 2 The present application provides a multi-modal depression disorder prediction system based on a viewing task, which comprises a physiological signal acquisition device, a computer, a viewing task module built in the computer, a feature extraction module, a data processing module, a prediction module and a display module. The viewing task module presents rest film materials, film materials for inducing acute psychological stress and film materials for restoring physical and mental state in turn according to a set time sequence. The present application sets three different presented film materials in the viewing task module, specifically including a rest viewing task unit, a stress viewing task unit and an intervention viewing task unit.

[0037] The first stage film materials presented in the rest viewing task unit are used for presenting to the viewer, and the viewing materials adopt calm and soothing rest film segments, which aims to make the viewer calm in the viewing process, and provide baseline collection for physiological data. The second stage film materials presented in the stress viewing task unit adopt material segments that can effectively induce acute psychological stress of the viewer, which provides a scene basis for testing the change of sympathetic nerve activity of the viewer through the change of individual physiological signals. The third stage film materials presented in the intervention viewing task unit adopt material segments that can effectively intervene and adjust the psychology of the viewer, which makes the viewer quickly restore the calm and harmony of the physical and mental state through beautiful visual appearance, soothing background music, soft voice guidance and other ways, and provides a scene basis for testing the change of parasympathetic nerve activity of the viewer through the change of individual physiological signals.

[0038] The physiological signal acquisition device is used for collecting pulse wave signals and skin resistance signals of the viewer during the viewing task; in the present application, the viewer wears a wearable physiological signal acquisition device integrated with a pulse wave (PPG) sensor and a skin resistance (GSR) sensor, and the physiological data of the person is synchronously acquired during the viewing task. Of course, two independent wearable devices can also be used, and the physiological signal acquisition device is connected to the computer, and the collected signals are transmitted to the computer.

[0039] The physiological signal acquisition device collects the skin resistance signal I of the viewer when performing the first stage of the resting viewing task, collects the pulse wave signal I and the skin resistance signal II of the viewer when performing the second stage of the stress viewing task, and collects the pulse wave signal II of the viewer when performing the third stage of the intervention viewing task.

[0040] The feature extraction module extracts the high-frequency power spectral density feature from the collected pulse wave signal and extracts the skin conductance level feature from the skin resistance signal, and transmits the two extracted features to the data processing module. Specifically, the skin conductance level feature I in the skin resistance signal I, the skin conductance level feature II in the skin resistance signal II, the high-frequency power spectral density feature I in the pulse wave signal I, and the high-frequency power spectral density feature II in the pulse wave signal II are extracted.

[0041] The data processing module calculates the feature difference value I of the skin conductance level and the feature difference value II of the high-frequency power spectral density during the viewing process of the viewer through the operation program. The data processing module first calculates the average value of the collected skin conductance level feature I, the skin conductance level feature II, the high-frequency power spectral density feature I and the high-frequency power spectral density feature II to obtain the resting skin resistance feature I, the stress skin resistance feature II, the stress heart rate variability high-frequency feature I and the intervention heart rate variability high-frequency feature II. The feature difference value I is the difference between the stress skin resistance feature II and the resting skin resistance feature I, and the feature difference value II is the difference between the intervention heart rate variability high-frequency feature II and the stress heart rate variability high-frequency feature I.

[0042] The prediction module is provided with a judgment threshold value I and a judgment threshold value II. The judgment threshold value I is preferably the standard deviation calculated by the data processing module for the skin conductance level feature I, and the judgment threshold value II is preferably the standard deviation calculated by the data processing module for the high-frequency power spectral density feature II. When the feature difference value I is higher than the judgment threshold value I and the feature difference value II is higher than the judgment threshold value II, it is predicted that the viewer has no risk of depression disorder, and the output is displayed on the computer through the display module.

[0043] The viewer is analyzed for depression disorder based on the above calculated values. When the feature difference value I of the viewer is higher than the judgment threshold value I, it is judged that the sympathetic nervous activity is normal, and a label A is given. When the feature difference value I is lower than the judgment threshold value I, it indicates that the individual cannot sensitively improve the sympathetic nervous activity in the stress-induced process, and it is judged that the sympathetic nervous activity is abnormal, and a label B is given.

[0044] When the feature difference II of the viewer is higher than the judgment threshold II, it is judged that the parasympathetic nerve activity is normal, and a label C is given; the feature difference II is lower than the judgment threshold II, indicating that the parasympathetic nerve regulation ability of the viewer is weak, and it is difficult to recover from the negative influence of stress, and it is judged that the parasympathetic nerve activity is abnormal, and a label D is given.

[0045] When the viewer has both A and C labels, the excitatory function of the sympathetic nerve and the regulation function of the parasympathetic nerve are normal, there is no risk of depression disorder, and output is outputted.

[0046] When the viewer has both B and C labels, or both A and D labels, the regulation function of the parasympathetic nerve is normal but the excitatory function of the sympathetic nerve is weak, or the excitatory function of the sympathetic nerve is normal but the regulation function of the parasympathetic nerve is weak, there is a risk of mild depression disorder, and output is outputted.

[0047] When the viewer has both B and D labels, the excitatory function of the sympathetic nerve and the regulation function of the parasympathetic nerve are both dysfunctional, there is a risk of severe depression disorder, and output is outputted.

[0048] Since autonomic dysfunction is a common feature of depression disorder, the sympathetic nervous system and the parasympathetic nervous system are two main branches of autonomic dysfunction. Individuals with depression disorder have poor autonomic activity, poor balance of sympathetic and parasympathetic nerves, and somatic symptoms caused by autonomic dysfunction imbalance are also the main factors for predicting depression disorder.

[0049] The present application integrates a pulse wave (PPG) sensor and a galvanic skin response (GSR) sensor into a wearable physiological signal acquisition device worn by a person, synchronously acquires physiological data of the person during the viewing task, and predicts the risk level of depression disorder of the individual through comprehensive analysis of the physiological signal changes of the individual during the viewing process.

[0050] The unmentioned parts of the present application are applicable to the prior art.

[0051] Obviously, the above embodiments are only examples for clearly illustrating, and are not limitations on the embodiments. For ordinary skilled persons in the art, other different forms of changes or variations can be made on the basis of the above description. Here, all the embodiments need not and cannot be exhausted. The obvious changes or variations derived therefrom are still within the protection scope of the present application.

Claims

1.A multi-modal depression disorder prediction system based on a viewing task, characterized in that, The system comprises: a viewing task module for sequentially presenting rest film materials, acute psychological stress inducing film materials and recovery of physical and mental state film materials in a set time sequence; a physiological signal acquisition device for acquiring pulse wave signals and skin resistance signals of a viewer during the execution of the viewing task; a feature extraction module for extracting high-frequency power spectral density features from the acquired pulse wave signals and skin conductance level features from the skin resistance signals; a data processing module with an operation program for calculating feature difference I of the skin conductance level and feature difference II of the high-frequency power spectral density of the viewer during the viewing process, respectively; a prediction module with a judgment threshold I and a judgment threshold II, when the feature difference I is higher than the judgment threshold I and the feature difference II is higher than the judgment threshold II, it is predicted that the viewer has no risk of depression disorder and the output is outputted, when at least one of the feature difference I and the feature difference II is lower than the corresponding judgment threshold I and the judgment threshold II, it is predicted that the viewer has a risk of depression disorder and the output is outputted; the viewing task module comprises: a rest viewing task unit containing calm and soothing rest film segments for enabling the viewer to achieve physical and mental calmness during the viewing process; a stress viewing task unit containing material segments for inducing acute psychological stress of the viewer for testing the change of sympathetic nerve activity of the viewer; an intervention viewing task unit containing material segments for psychological intervention and adjustment of the viewer for enabling the viewer to quickly recover the calmness and harmony of the physical and mental state and testing the change of parasympathetic nerve activity of the viewer; the physiological signal acquisition device acquires the skin resistance signal I of the viewer during the execution of the rest viewing task, the pulse wave signal I and the skin resistance signal II of the viewer during the execution of the stress viewing task, and the pulse wave signal II of the viewer during the execution of the intervention viewing task; the feature extraction module sequentially extracts the skin conductance level feature I in the skin resistance signal I, the skin conductance level feature II in the skin resistance signal II, the high-frequency power spectral density feature I in the pulse wave signal I and the high-frequency power spectral density feature II in the pulse wave signal II. 2.The movie watching task based multi-modal depression disorder prediction system according to claim 1, wherein, The data processing module calculates the average values of the acquired skin conductance level features I, the skin conductance level features II, the high-frequency power spectral density features I and the high-frequency power spectral density features II, respectively, to obtain the rest skin resistance feature I, the stress skin resistance feature II, the stress heart rate variability high-frequency feature I and the intervention heart rate variability high-frequency feature II; the feature difference I is the difference between the stress skin resistance feature II and the rest skin resistance feature I, and the feature difference II is the difference between the intervention heart rate variability high-frequency feature II and the stress heart rate variability high-frequency feature I. 3.The movie watching task based multi-modal depression disorder prediction system according to claim 2, wherein, The judgment threshold I in the prediction module is the standard deviation calculated by the data processing module for the skin conductance level feature I, and the judgment threshold II is the standard deviation calculated by the data processing module for the high-frequency power spectral density feature II. 4.The movie watching task based multi-modal depression disorder prediction system according to claim 3, wherein, The prediction module makes a grading prediction on the viewer with a risk of depression disorder: when the feature difference I is higher than the judgment threshold I, the viewer is judged to have normal sympathetic nervous activity, and label A is output; otherwise, the viewer is judged to have sympathetic nervous activity disorder, and label B is output; When the feature difference II is higher than the judgment threshold II, the viewer is judged to have normal parasympathetic nervous activity, and label C is output; otherwise, the viewer is judged to have parasympathetic nervous activity disorder, and label D is output. 5.The movie-watching task based multi-modal depression disorder prediction system according to claim 4, wherein, The prediction module also makes a grading prediction on the output label of depression disorder: when the viewer gets label A and label C at the same time, the viewer is judged to have no risk of depression disorder, and is output; When the viewer gets label B and label D at the same time, the viewer is judged to have a risk of severe depression disorder, and is output; When the viewer gets label A and label D, or label B and label C at the same time, the viewer is judged to have a risk of mild depression disorder, and is output. 6.The movie-watching task based multi-modal depression disorder prediction system of claim 1, wherein, The physiological signal acquisition device is a wearable pulse wave sensor and a skin resistance sensor, or a wearable acquisition device integrated with a pulse wave sensor and a skin resistance sensor.

Images