Pupil wave-based Hamilton depression scale evaluation method and system

Abstract

The invention discloses a Hamilton depression scale evaluation method and system based on pupil waves, and relates to the technical field of emotional psychology, artificial intelligence and deep learning, and the system comprises an emotion video display module, a pupil wave generation module, a data preprocessing module and a scale evaluation module. The method comprises the following steps: firstly, selecting a positive video and a negative video from an emotion database through an emotion video display module, and displaying and playing the positive video and the negative video for an assessed person to watch; secondly, pupil waves are generated according to the eye image through a pupil wave generation module; then, the collected pupil waves are preprocessed through a data preprocessing module; then, through a scale evaluation module, a Hamilton scale value is evaluated according to an evaluation model. Therefore, by adopting the pupil wave-based Hamilton depression scale evaluation method and system, the Hamilton scale value can be measured and evaluated more objectively and efficiently, and a scientific basis is provided for clinical depression diagnosis and curative effect evaluation.

Description

Technical Field

[0001] The present invention relates to the technical fields of emotion psychology, artificial intelligence, and deep learning, and particularly relates to a method and system for assessing the Hamilton Depression Scale based on pupil waves. Background Art

[0002] Currently, clinical diagnosis of depression mainly relies on the experience of psychiatrists and scale information, especially the Hamilton Depression Scale information.

[0003] The assessment of Hamilton scale values is generally carried out by two doctors or two trained professionals, with relatively high accuracy. However, the manual assessment of Hamilton is time-consuming and laborious, and it is difficult to form an efficient assessment standard and process. Therefore, it is necessary to provide an efficient and accurate assessment method to provide a scientific basis for clinical depression diagnosis.

[0004] The eyes are the window to the soul and an important signal for people's emotional communication. Some studies have shown that patients with depression have insufficient happy experiences, excessive sad or negative emotional experiences. Compared with non-depressed patients, their pupils dilate when watching external happy emotions and constrict when watching sad or negative emotions. Therefore, using pupil wave signals for automatic assessment of Hamilton scale values is expected to provide an objective and highly accurate assessment scheme. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for assessing the Hamilton Depression Scale based on pupil waves, which can effectively extract the characteristics of pupil wave signals and achieve objective and highly accurate automatic assessment of Hamilton scale values.

[0006] To achieve the above purpose, the present invention provides a method for assessing the Hamilton Depression Scale based on pupil waves, including the following steps:

[0007] First, through an emotion video display module, select negative videos and positive videos from a preset emotion database, and display and play them for the assessor to watch;

[0008] Second, through a pupil wave generation module, collect the eye images of the assessor when watching the video and generate pupil waves;

[0009] Next, through a data preprocessing module, perform outlier processing and noise reduction processing on the collected pupil waves;

[0010] Then, through a scale assessment module, train the scale assessment model to obtain the Hamilton scale value.

[0011] Preferably, generating pupil waves includes calculating the pupil diameter based on the eye image and forming the calculated data of the time series into pupil waves.

[0012] Preferably, the outlier processing includes smoothing the pupil wave values less than the threshold when the eyes are closed or semi-closed, as follows:

[0013]

[0014] In the formula, is the replacement value of the abnormal pupil diameter P at the t-th moment of the pupil wave sequence, and P t 、P t-1 、P t+1 are the pupil diameters on the left and right sides at the t-th moment, respectively.

[0015] Preferably, the denoising process is carried out by a combined filtering method of wavelet filtering and Kalman filtering, including removing noise by the wavelet filtering method and then removing the remaining noise by the Kalman filtering method.

[0016] Preferably, training the scale assessment model includes:

[0017] First, obtain the pupil wave sample data corresponding to the individual;

[0018] Secondly, divide the sample data into a training set, a validation set, and a test set;

[0019] Then, use the training set to train the deep learning model, use the validation set to validate the deep learning model, use the test set to test the deep learning model, and use the trained deep learning model as the scale assessment model.

[0020] A Hamilton depression scale assessment system based on pupil waves includes:

[0021] An emotional video display module for selecting negative and positive videos and displaying and playing them for the assessor to watch;

[0022] A pupil wave generation module for collecting eye images and generating pupil waves;

[0023] A data preprocessing module for preprocessing the collected pupil waves;

[0024] A scale assessment module for obtaining a scale assessment model and assessing the Hamilton scale value.

[0025] Preferably, the emotional video display module includes an emotional video selection, a video clip, and a video display part;

[0026] Among them, the emotional video selection part is used to select a sad emotional video and a happy emotional video from the emotional video library;

[0027] The video clip part is used to crop the emotional video to a fixed duration;

[0028] The video display part sequentially plays sad and happy videos for the assessors through a monitor.

[0029] Therefore, the present invention adopts the above-mentioned method and system for assessing the Hamilton Depression Rating Scale based on pupil waves, and has the following technical effects:

[0030] (1) Using pupil waves to assess the Hamilton Depression Rating Scale. As a physiological index, the change in pupil size is closely related to emotions. At the physiological level, the direct manifestation is that the visual stimulation of happy emotional information will cause the subjects to have a pleasant emotional experience, which in turn leads to an increase in pupil diameter; the visual stimulation of sad or negative emotional information will cause the subjects to have a sad emotional experience, directly manifested as a decrease in pupil diameter. Compared with physiological signals such as functional magnetic resonance brain imaging or brain waves, and behavioral signals such as facial expressions, speech, and language, pupil waves can more directly reflect the most instinctive and unmaskable emotional expressions.

[0031] (2) Providing a system that integrates data collection, processing, and assessment of the Hamilton Depression Rating Scale, realizing the efficient and high-precision assessment of the Hamilton Depression Rating Scale.

[0032] The following will further describe the technical solutions of the present invention in detail through the drawings and embodiments. Description of the Drawings

[0033] Figure 1 is a schematic diagram of a system for assessing the Hamilton Depression Rating Scale based on pupil waves;

[0034] Figure 2 is a schematic diagram of an emotional video display module in a system for assessing the Hamilton Depression Rating Scale based on pupil waves;

[0035] Figure 3 is a schematic diagram of a pupil wave generation module in a system for assessing the Hamilton Depression Rating Scale based on pupil waves;

[0036] Figure 4 is a schematic diagram of pupil wave preprocessing in an embodiment of a method and system for assessing the Hamilton Depression Rating Scale based on pupil waves;

[0037] Figure 5 is a schematic diagram of assessment model training in an embodiment of a method and system for assessing the Hamilton Depression Rating Scale based on pupil waves. Detailed Embodiments

[0038] The present invention can be more detailedly explained through the following embodiments. The purpose of disclosing the present invention is to protect all changes and improvements within the scope of the present invention. The present invention is not limited to the following embodiments.

[0039] As Figure 1As shown in the figure, the present invention provides a Hamilton depression scale assessment system based on pupil waves, including an emotional video display module for selecting negative and positive videos and displaying and playing them for the assessor to watch; a pupil wave generation module for collecting eye images and generating pupil waves; a data preprocessing module for preprocessing the collected pupil waves; and a scale assessment module for obtaining a scale assessment model and assessing the Hamilton scale value.

[0040] Among them, the emotional video display module includes an emotional video selection, video editing, and video display part: the emotional video selection part is used to select a sad emotional video and a happy emotional video from the emotional video library; the video editing part is used to crop the emotional video to a fixed duration; and the video display part sequentially plays the sad and happy videos for the assessor through a display.

[0041] The present invention also provides a method for assessing the Hamilton depression scale based on pupil waves, including the following steps:

[0042] As Figure 2 shown, through the emotional video display module, a sad emotional video is randomly selected from the negative emotional video library, and a happy emotional video is randomly selected from the positive emotional video library; the sad emotional video and the happy emotional video are respectively edited into videos with a duration of 90 seconds, and the sad and happy videos are sequentially played for the assessor through a display.

[0043] As Figure 3 shown, through the pupil wave generation module, the eye images of 65 assessors are synchronously collected using a camera while they are watching the emotional videos. Specifically, the frequency of the pupil wave acquisition camera is 60 frames per second, and the pixels are 800×600. Then, using the pupil wave generation software, the pupil diameter of each frame of the eye image is calculated in real time, and its time series forms a pupil wave.

[0044] As Figure 4 shown, through the data preprocessing module, outlier processing and noise removal processing are performed on the collected pupil waves, specifically as follows:

[0045] For outlier processing, the pupil wave values less than the threshold when the eyes are closed or semi-closed are replaced through the mean smoothing processing method, and the expression is as follows:

[0046]

[0047] In the formula, is the replacement value of the pupil diameter outlier P t at the t-th moment of the pupil wave sequence, and P t-1 , P t+1 are the pupil diameters on the left and right sides at the t-th moment, respectively.

[0048] For noise reduction processing, a combined filtering method of wavelet filtering and Kalman filtering is adopted. The wavelet filtering method is used to remove noise, and the Kalman filtering method is used to remove the remaining noise.

[0049] As Figure 5 shown, through the scale evaluation module, the scale evaluation model is trained to obtain the Hamilton scale value, specifically as follows:

[0050] First, obtain the pupil wave sample data corresponding to multiple individuals, a total of 65 sample data.

[0051] Second, divide the 65 sample data into a training set, a validation set, and a test set according to a ratio of 6:2:2, where the training set:validation set:test set = 39:13:13.

[0052] Then, use the training set to train the deep learning model, use the validation set to validate the deep learning model, use the test set to test the deep learning model, and use the trained deep learning model as the scale evaluation model.

[0053] Finally, input the pupil wave signal of the person to be evaluated into the evaluation model to obtain the Hamilton scale evaluation value.

[0054] Therefore, the present invention adopts the above-mentioned method and system for evaluating the Hamilton depression scale based on pupil waves, which can provide an objective standard for evaluating the Hamilton scale score, and evaluate the Hamilton scale score through pupil wave signals and deep learning technology, realizing efficient and high-precision automatic evaluation of the Hamilton scale.

[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that they can still modify or equivalently replace the technical solutions of the present invention, and these modifications or equivalent replacements cannot make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A pupil wave-based Hamilton Depression Rating Scale assessment method, characterized in that: The following steps are involved: First, through the emotional video display module, negative videos and positive videos are selected from the preset emotional database, and displayed and played for the evaluated person to watch; Secondly, through the pupil wave generation module, the eye images of the person being evaluated when watching the video are collected and the pupil waves are generated; Next, the collected pupil waves are processed for outlier value and noise removal through the data preprocessing module; Then, the scale assessment model is trained through the scale assessment module to obtain the Hamilton scale value.

2. The Hamilton Depression Rating Scale assessment method based on pupil waves according to claim 1, characterized in that: Generating the pupil wave includes calculating the pupil diameter according to the eye image and composing the time series calculated data into the pupil wave.

3. The Hamilton Depression Rating Scale assessment method based on pupil waves according to claim 1, characterized in that: Outlier processing includes smoothing the pupil wave values ​​that are less than the threshold when the eyes are closed or half closed, as follows: In the formula, is the abnormal value of pupil diameter P at time t of pupil wave sequence t The replacement value, P t-1 , P t+1 are the left and right pupil diameters at time t respectively.

4. The Hamilton Depression Rating Scale assessment method based on pupil waves according to claim 1, characterized in that: The denoising process uses a combined filtering method of wavelet filtering and Kalman filtering, including using a wavelet filtering method to remove noise and then using a Kalman filtering method to remove the remaining noise.

5. The Hamilton Depression Rating Scale assessment method based on pupil waves according to claim 1, characterized in that: Training the scale assessment model includes: First, obtain the pupil wave sample data corresponding to the individual; Secondly, divide the sample data into training set, validation set and test set; Then, the training set is used to train the deep learning model, the validation set is used to validate the deep learning model, the test set is used to test the deep learning model, and the trained deep learning model is used as the scale evaluation model.

6. A Hamilton Depression Rating Scale assessment system based on pupil waves, characterized in that: include: The emotional video display module is used to select negative videos and positive videos and display them for the person being evaluated to watch; A pupil wave generation module, used for collecting eye images and generating pupil waves; A data preprocessing module, used for preprocessing the collected pupil waves; The scale assessment module is used to obtain the scale assessment model and assess the Hamilton scale value.

7. The Hamilton Depression Rating Scale assessment system based on pupil waves according to claim 6, characterized in that: The emotional video display module includes emotional video selection, video editing and video display parts; The emotional video selection part is used to select a sad emotional video and a happy emotional video from the emotional video library; The video clipping part is used to cut the emotional video into a fixed length; In the video display part, sad and happy videos are played to the person being assessed in sequence through a display.

Images