Cargando…
Detecting Depression Through Gait Data: Examining the Contribution of Gait Features in Recognizing Depression
While depression is one of the most common mental disorders affecting more than 300 million people across the world, it is often left undiagnosed. This paper investigated the association between depression and gait characteristics with the aim to assist in diagnosing depression. Our dataset consiste...
Autores principales: | , , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
Frontiers Media S.A.
2021
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8138135/ https://www.ncbi.nlm.nih.gov/pubmed/34025483 http://dx.doi.org/10.3389/fpsyt.2021.661213 |
_version_ | 1783695750532694016 |
---|---|
author | Wang, Yameng Wang, Jingying Liu, Xiaoqian Zhu, Tingshao |
author_facet | Wang, Yameng Wang, Jingying Liu, Xiaoqian Zhu, Tingshao |
author_sort | Wang, Yameng |
collection | PubMed |
description | While depression is one of the most common mental disorders affecting more than 300 million people across the world, it is often left undiagnosed. This paper investigated the association between depression and gait characteristics with the aim to assist in diagnosing depression. Our dataset consisted of 121 healthy people and 126 patients with depression who diagnosed by psychiatrists according to the Diagnostic and Statistical Manual of Mental Disorders. Spatiotemporal, temporal-domain, and frequency-domain features were extracted based on the walking data of 247 participants recorded by Microsoft Kinect (Version 2). Multiple logistic regression was used to analyze the variance of spatiotemporal (12.55%), time-domain (58.36%), and frequency-domain features (60.71%) on recognizing depression based on Nagelkerke's R(2) measure, respectively. The contributions of the different types of features were further explored by building machine learning models by using support vector machine algorithm. All the combinations of the three types of gait features were used as training data of machine learning models, respectively. The results showed that the model trained using only time- and frequency-domain features demonstrated the same best performance compared to the model trained using all the features (sensitivity = 0.94, specificity = 0.91, and AUC = 0.93). These results indicated that depression could be effectively recognized through gait analysis. This approach is a step forward toward developing low-cost, non-intrusive solutions for real-time depression recognition. |
format | Online Article Text |
id | pubmed-8138135 |
institution | National Center for Biotechnology Information |
language | English |
publishDate | 2021 |
publisher | Frontiers Media S.A. |
record_format | MEDLINE/PubMed |
spelling | pubmed-81381352021-05-22 Detecting Depression Through Gait Data: Examining the Contribution of Gait Features in Recognizing Depression Wang, Yameng Wang, Jingying Liu, Xiaoqian Zhu, Tingshao Front Psychiatry Psychiatry While depression is one of the most common mental disorders affecting more than 300 million people across the world, it is often left undiagnosed. This paper investigated the association between depression and gait characteristics with the aim to assist in diagnosing depression. Our dataset consisted of 121 healthy people and 126 patients with depression who diagnosed by psychiatrists according to the Diagnostic and Statistical Manual of Mental Disorders. Spatiotemporal, temporal-domain, and frequency-domain features were extracted based on the walking data of 247 participants recorded by Microsoft Kinect (Version 2). Multiple logistic regression was used to analyze the variance of spatiotemporal (12.55%), time-domain (58.36%), and frequency-domain features (60.71%) on recognizing depression based on Nagelkerke's R(2) measure, respectively. The contributions of the different types of features were further explored by building machine learning models by using support vector machine algorithm. All the combinations of the three types of gait features were used as training data of machine learning models, respectively. The results showed that the model trained using only time- and frequency-domain features demonstrated the same best performance compared to the model trained using all the features (sensitivity = 0.94, specificity = 0.91, and AUC = 0.93). These results indicated that depression could be effectively recognized through gait analysis. This approach is a step forward toward developing low-cost, non-intrusive solutions for real-time depression recognition. Frontiers Media S.A. 2021-05-07 /pmc/articles/PMC8138135/ /pubmed/34025483 http://dx.doi.org/10.3389/fpsyt.2021.661213 Text en Copyright © 2021 Wang, Wang, Liu and Zhu. https://creativecommons.org/licenses/by/4.0/This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. |
spellingShingle | Psychiatry Wang, Yameng Wang, Jingying Liu, Xiaoqian Zhu, Tingshao Detecting Depression Through Gait Data: Examining the Contribution of Gait Features in Recognizing Depression |
title | Detecting Depression Through Gait Data: Examining the Contribution of Gait Features in Recognizing Depression |
title_full | Detecting Depression Through Gait Data: Examining the Contribution of Gait Features in Recognizing Depression |
title_fullStr | Detecting Depression Through Gait Data: Examining the Contribution of Gait Features in Recognizing Depression |
title_full_unstemmed | Detecting Depression Through Gait Data: Examining the Contribution of Gait Features in Recognizing Depression |
title_short | Detecting Depression Through Gait Data: Examining the Contribution of Gait Features in Recognizing Depression |
title_sort | detecting depression through gait data: examining the contribution of gait features in recognizing depression |
topic | Psychiatry |
url | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8138135/ https://www.ncbi.nlm.nih.gov/pubmed/34025483 http://dx.doi.org/10.3389/fpsyt.2021.661213 |
work_keys_str_mv | AT wangyameng detectingdepressionthroughgaitdataexaminingthecontributionofgaitfeaturesinrecognizingdepression AT wangjingying detectingdepressionthroughgaitdataexaminingthecontributionofgaitfeaturesinrecognizingdepression AT liuxiaoqian detectingdepressionthroughgaitdataexaminingthecontributionofgaitfeaturesinrecognizingdepression AT zhutingshao detectingdepressionthroughgaitdataexaminingthecontributionofgaitfeaturesinrecognizingdepression |