Using 2D video-based pose estimation for automated prediction of autism spectrum disorders in young children

Abstract Clinical research in autism has recently witnessed promising digital phenotyping results, mainly focused on single feature extraction, such as gaze, head turn on name-calling or visual tracking of the moving object. The main drawback of these studies is the focus on relatively isolated beha...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Nada Kojovic, Shreyasvi Natraj, Sharada Prasanna Mohanty, Thomas Maillart, Marie Schaer
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2021
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/f0ec9e65eaf94a928d1f9dddd7e118e9
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:f0ec9e65eaf94a928d1f9dddd7e118e9
record_format	dspace
spelling	oai:doaj.org-article:f0ec9e65eaf94a928d1f9dddd7e118e92021-12-02T16:50:24ZUsing 2D video-based pose estimation for automated prediction of autism spectrum disorders in young children10.1038/s41598-021-94378-z2045-2322https://doaj.org/article/f0ec9e65eaf94a928d1f9dddd7e118e92021-07-01T00:00:00Zhttps://doi.org/10.1038/s41598-021-94378-zhttps://doaj.org/toc/2045-2322Abstract Clinical research in autism has recently witnessed promising digital phenotyping results, mainly focused on single feature extraction, such as gaze, head turn on name-calling or visual tracking of the moving object. The main drawback of these studies is the focus on relatively isolated behaviors elicited by largely controlled prompts. We recognize that while the diagnosis process understands the indexing of the specific behaviors, ASD also comes with broad impairments that often transcend single behavioral acts. For instance, the atypical nonverbal behaviors manifest through global patterns of atypical postures and movements, fewer gestures used and often decoupled from visual contact, facial affect, speech. Here, we tested the hypothesis that a deep neural network trained on the non-verbal aspects of social interaction can effectively differentiate between children with ASD and their typically developing peers. Our model achieves an accuracy of 80.9% (F1 score: 0.818; precision: 0.784; recall: 0.854) with the prediction probability positively correlated to the overall level of symptoms of autism in social affect and repetitive and restricted behaviors domain. Provided the non-invasive and affordable nature of computer vision, our approach carries reasonable promises that a reliable machine-learning-based ASD screening may become a reality not too far in the future.Nada KojovicShreyasvi NatrajSharada Prasanna MohantyThomas MaillartMarie SchaerNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 11, Iss 1, Pp 1-10 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Nada Kojovic Shreyasvi Natraj Sharada Prasanna Mohanty Thomas Maillart Marie Schaer Using 2D video-based pose estimation for automated prediction of autism spectrum disorders in young children
description	Abstract Clinical research in autism has recently witnessed promising digital phenotyping results, mainly focused on single feature extraction, such as gaze, head turn on name-calling or visual tracking of the moving object. The main drawback of these studies is the focus on relatively isolated behaviors elicited by largely controlled prompts. We recognize that while the diagnosis process understands the indexing of the specific behaviors, ASD also comes with broad impairments that often transcend single behavioral acts. For instance, the atypical nonverbal behaviors manifest through global patterns of atypical postures and movements, fewer gestures used and often decoupled from visual contact, facial affect, speech. Here, we tested the hypothesis that a deep neural network trained on the non-verbal aspects of social interaction can effectively differentiate between children with ASD and their typically developing peers. Our model achieves an accuracy of 80.9% (F1 score: 0.818; precision: 0.784; recall: 0.854) with the prediction probability positively correlated to the overall level of symptoms of autism in social affect and repetitive and restricted behaviors domain. Provided the non-invasive and affordable nature of computer vision, our approach carries reasonable promises that a reliable machine-learning-based ASD screening may become a reality not too far in the future.
format	article
author	Nada Kojovic Shreyasvi Natraj Sharada Prasanna Mohanty Thomas Maillart Marie Schaer
author_facet	Nada Kojovic Shreyasvi Natraj Sharada Prasanna Mohanty Thomas Maillart Marie Schaer
author_sort	Nada Kojovic
title	Using 2D video-based pose estimation for automated prediction of autism spectrum disorders in young children
title_short	Using 2D video-based pose estimation for automated prediction of autism spectrum disorders in young children
title_full	Using 2D video-based pose estimation for automated prediction of autism spectrum disorders in young children
title_fullStr	Using 2D video-based pose estimation for automated prediction of autism spectrum disorders in young children
title_full_unstemmed	Using 2D video-based pose estimation for automated prediction of autism spectrum disorders in young children
title_sort	using 2d video-based pose estimation for automated prediction of autism spectrum disorders in young children
publisher	Nature Portfolio
publishDate	2021
url	https://doaj.org/article/f0ec9e65eaf94a928d1f9dddd7e118e9
work_keys_str_mv	AT nadakojovic using2dvideobasedposeestimationforautomatedpredictionofautismspectrumdisordersinyoungchildren AT shreyasvinatraj using2dvideobasedposeestimationforautomatedpredictionofautismspectrumdisordersinyoungchildren AT sharadaprasannamohanty using2dvideobasedposeestimationforautomatedpredictionofautismspectrumdisordersinyoungchildren AT thomasmaillart using2dvideobasedposeestimationforautomatedpredictionofautismspectrumdisordersinyoungchildren AT marieschaer using2dvideobasedposeestimationforautomatedpredictionofautismspectrumdisordersinyoungchildren
_version_	1718383074428846080

Using 2D video-based pose estimation for automated prediction of autism spectrum disorders in young children

Ejemplares similares