A New Regression Model for Depression Severity Prediction Based on Correlation among Audio Features Using a Graph Convolutional Neural Network
Recent studies have revealed mutually correlated audio features in the voices of depressed patients. Thus, the voices of these patients can be characterized based on the combinatorial relationships among the audio features. To date, many deep learning–based methods have been proposed to predict the...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-02-01
|
| Series: | Diagnostics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-4418/13/4/727 |
| _version_ | 1827757824405405696 |
|---|---|
| author | Momoko Ishimaru Yoshifumi Okada Ryunosuke Uchiyama Ryo Horiguchi Itsuki Toyoshima |
| author_facet | Momoko Ishimaru Yoshifumi Okada Ryunosuke Uchiyama Ryo Horiguchi Itsuki Toyoshima |
| author_sort | Momoko Ishimaru |
| collection | DOAJ |
| description | Recent studies have revealed mutually correlated audio features in the voices of depressed patients. Thus, the voices of these patients can be characterized based on the combinatorial relationships among the audio features. To date, many deep learning–based methods have been proposed to predict the depression severity using audio data. However, existing methods have assumed that the individual audio features are independent. Hence, in this paper, we propose a new deep learning–based regression model that allows for the prediction of depression severity on the basis of the correlation among audio features. The proposed model was developed using a graph convolutional neural network. This model trains the voice characteristics using graph-structured data generated to express the correlation among audio features. We conducted prediction experiments on depression severity using the DAIC-WOZ dataset employed in several previous studies. The experimental results showed that the proposed model achieved a root mean square error (RMSE) of 2.15, a mean absolute error (MAE) of 1.25, and a symmetric mean absolute percentage error of 50.96%. Notably, RMSE and MAE significantly outperformed the existing state-of-the-art prediction methods. From these results, we conclude that the proposed model can be a promising tool for depression diagnosis. |
| first_indexed | 2024-03-11T08:56:57Z |
| format | Article |
| id | doaj.art-67ff5917e0c3475ca10a07f3d10e61d5 |
| institution | Directory Open Access Journal |
| issn | 2075-4418 |
| language | English |
| last_indexed | 2024-03-11T08:56:57Z |
| publishDate | 2023-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Diagnostics |
| spelling | doaj.art-67ff5917e0c3475ca10a07f3d10e61d52023-11-16T20:02:03ZengMDPI AGDiagnostics2075-44182023-02-0113472710.3390/diagnostics13040727A New Regression Model for Depression Severity Prediction Based on Correlation among Audio Features Using a Graph Convolutional Neural NetworkMomoko Ishimaru0Yoshifumi Okada1Ryunosuke Uchiyama2Ryo Horiguchi3Itsuki Toyoshima4Division of Information and Electronic Engineering, Muroran Institute of Technology, 27-1, Mizumoto-cho, Muroran 050-8585, JapanCollege of Information and Systems, Muroran Institute of Technology, 27-1, Mizumoto-cho, Muroran 050-8585, JapanDivision of Information and Electronic Engineering, Muroran Institute of Technology, 27-1, Mizumoto-cho, Muroran 050-8585, JapanDivision of Information and Electronic Engineering, Muroran Institute of Technology, 27-1, Mizumoto-cho, Muroran 050-8585, JapanDivision of Information and Electronic Engineering, Muroran Institute of Technology, 27-1, Mizumoto-cho, Muroran 050-8585, JapanRecent studies have revealed mutually correlated audio features in the voices of depressed patients. Thus, the voices of these patients can be characterized based on the combinatorial relationships among the audio features. To date, many deep learning–based methods have been proposed to predict the depression severity using audio data. However, existing methods have assumed that the individual audio features are independent. Hence, in this paper, we propose a new deep learning–based regression model that allows for the prediction of depression severity on the basis of the correlation among audio features. The proposed model was developed using a graph convolutional neural network. This model trains the voice characteristics using graph-structured data generated to express the correlation among audio features. We conducted prediction experiments on depression severity using the DAIC-WOZ dataset employed in several previous studies. The experimental results showed that the proposed model achieved a root mean square error (RMSE) of 2.15, a mean absolute error (MAE) of 1.25, and a symmetric mean absolute percentage error of 50.96%. Notably, RMSE and MAE significantly outperformed the existing state-of-the-art prediction methods. From these results, we conclude that the proposed model can be a promising tool for depression diagnosis.https://www.mdpi.com/2075-4418/13/4/727audio featuredepressionregression modelcorrelationgraph convolutional neural network |
| spellingShingle | Momoko Ishimaru Yoshifumi Okada Ryunosuke Uchiyama Ryo Horiguchi Itsuki Toyoshima A New Regression Model for Depression Severity Prediction Based on Correlation among Audio Features Using a Graph Convolutional Neural Network Diagnostics audio feature depression regression model correlation graph convolutional neural network |
| title | A New Regression Model for Depression Severity Prediction Based on Correlation among Audio Features Using a Graph Convolutional Neural Network |
| title_full | A New Regression Model for Depression Severity Prediction Based on Correlation among Audio Features Using a Graph Convolutional Neural Network |
| title_fullStr | A New Regression Model for Depression Severity Prediction Based on Correlation among Audio Features Using a Graph Convolutional Neural Network |
| title_full_unstemmed | A New Regression Model for Depression Severity Prediction Based on Correlation among Audio Features Using a Graph Convolutional Neural Network |
| title_short | A New Regression Model for Depression Severity Prediction Based on Correlation among Audio Features Using a Graph Convolutional Neural Network |
| title_sort | new regression model for depression severity prediction based on correlation among audio features using a graph convolutional neural network |
| topic | audio feature depression regression model correlation graph convolutional neural network |
| url | https://www.mdpi.com/2075-4418/13/4/727 |
| work_keys_str_mv | AT momokoishimaru anewregressionmodelfordepressionseveritypredictionbasedoncorrelationamongaudiofeaturesusingagraphconvolutionalneuralnetwork AT yoshifumiokada anewregressionmodelfordepressionseveritypredictionbasedoncorrelationamongaudiofeaturesusingagraphconvolutionalneuralnetwork AT ryunosukeuchiyama anewregressionmodelfordepressionseveritypredictionbasedoncorrelationamongaudiofeaturesusingagraphconvolutionalneuralnetwork AT ryohoriguchi anewregressionmodelfordepressionseveritypredictionbasedoncorrelationamongaudiofeaturesusingagraphconvolutionalneuralnetwork AT itsukitoyoshima anewregressionmodelfordepressionseveritypredictionbasedoncorrelationamongaudiofeaturesusingagraphconvolutionalneuralnetwork AT momokoishimaru newregressionmodelfordepressionseveritypredictionbasedoncorrelationamongaudiofeaturesusingagraphconvolutionalneuralnetwork AT yoshifumiokada newregressionmodelfordepressionseveritypredictionbasedoncorrelationamongaudiofeaturesusingagraphconvolutionalneuralnetwork AT ryunosukeuchiyama newregressionmodelfordepressionseveritypredictionbasedoncorrelationamongaudiofeaturesusingagraphconvolutionalneuralnetwork AT ryohoriguchi newregressionmodelfordepressionseveritypredictionbasedoncorrelationamongaudiofeaturesusingagraphconvolutionalneuralnetwork AT itsukitoyoshima newregressionmodelfordepressionseveritypredictionbasedoncorrelationamongaudiofeaturesusingagraphconvolutionalneuralnetwork |