Automated pose estimation captures key aspects of General Movements at 8-17 weeks from conventional videos

Viviana Marchi, Anna Hakala, Andrew Knight, Federica D'Acunto, Maria Luisa Scattoni, Andrea Guzzetta, Sampsa Vanhatalo

Forskningsoutput: TidskriftsbidragArtikelVetenskapligPeer review

Sammanfattning

Abstract Aim General movement assessment requires substantial expertise for accurate visual interpretation. Our aim was to evaluate an automated pose estimation method, using conventional video records, to see if it could capture infant movements using objective biomarkers. Methods We selected archived videos from 21 infants aged 8-17 weeks who had taken part in studies at the IRCCS Stella Maris Foundation (Italy), from 2011-2017. Of these, 14 presented with typical low-risk movements, while seven presented with atypical movements and were later diagnosed with cerebral palsy. Skeleton videos were produced using a computational pose estimation model adapted for infants and these were blindly assessed to see whether they contained the information needed for classification by human experts. Movements of skeletal key points were analysed using kinematic metrics to provide a biomarker to distinguish between groups. Results The visual assessments of the skeleton videos were very accurate, with Cohen's K of 0.90 when compared with the classification of conventional videos. Quantitative analysis showed that arm movements were more variable in infants with typical movements. Conclusion It was possible to extract automated estimation of movement patterns from conventional video records and convert them to skeleton footage. This could allow quantitative analysis of existing footage. This article is protected by copyright. All rights reserved.
Originalspråkengelska
TidskriftActa Paediatrica
Volym0
Utgåvaja
ISSN0803-5253
DOI
Status!!E-pub ahead of print - 18 mar 2019
MoE-publikationstypA1 Tidskriftsartikel-refererad

Vetenskapsgrenar

  • 3123 Kvinno- och barnsjukdomar

Citera det här

Marchi, Viviana ; Hakala, Anna ; Knight, Andrew ; D'Acunto, Federica ; Scattoni, Maria Luisa ; Guzzetta, Andrea ; Vanhatalo, Sampsa. / Automated pose estimation captures key aspects of General Movements at 8-17 weeks from conventional videos. I: Acta Paediatrica. 2019 ; Vol. 0, Nr. ja.
@article{72118f34b9ef4b958838269007718ad5,
title = "Automated pose estimation captures key aspects of General Movements at 8-17 weeks from conventional videos",
abstract = "Abstract Aim General movement assessment requires substantial expertise for accurate visual interpretation. Our aim was to evaluate an automated pose estimation method, using conventional video records, to see if it could capture infant movements using objective biomarkers. Methods We selected archived videos from 21 infants aged 8-17 weeks who had taken part in studies at the IRCCS Stella Maris Foundation (Italy), from 2011-2017. Of these, 14 presented with typical low-risk movements, while seven presented with atypical movements and were later diagnosed with cerebral palsy. Skeleton videos were produced using a computational pose estimation model adapted for infants and these were blindly assessed to see whether they contained the information needed for classification by human experts. Movements of skeletal key points were analysed using kinematic metrics to provide a biomarker to distinguish between groups. Results The visual assessments of the skeleton videos were very accurate, with Cohen's K of 0.90 when compared with the classification of conventional videos. Quantitative analysis showed that arm movements were more variable in infants with typical movements. Conclusion It was possible to extract automated estimation of movement patterns from conventional video records and convert them to skeleton footage. This could allow quantitative analysis of existing footage. This article is protected by copyright. All rights reserved.",
keywords = "automatic pose estimation, cerebral palsy, general movements, infants, spontaneous motor activity, 3123 Gynaecology and paediatrics",
author = "Viviana Marchi and Anna Hakala and Andrew Knight and Federica D'Acunto and Scattoni, {Maria Luisa} and Andrea Guzzetta and Sampsa Vanhatalo",
year = "2019",
month = "3",
day = "18",
doi = "10.1111/apa.14781",
language = "English",
volume = "0",
journal = "Acta Paediatrica",
issn = "0803-5253",
publisher = "Wiley",
number = "ja",

}

Automated pose estimation captures key aspects of General Movements at 8-17 weeks from conventional videos. / Marchi, Viviana; Hakala, Anna; Knight, Andrew; D'Acunto, Federica; Scattoni, Maria Luisa; Guzzetta, Andrea; Vanhatalo, Sampsa.

I: Acta Paediatrica, Vol. 0, Nr. ja, 18.03.2019.

Forskningsoutput: TidskriftsbidragArtikelVetenskapligPeer review

TY - JOUR

T1 - Automated pose estimation captures key aspects of General Movements at 8-17 weeks from conventional videos

AU - Marchi, Viviana

AU - Hakala, Anna

AU - Knight, Andrew

AU - D'Acunto, Federica

AU - Scattoni, Maria Luisa

AU - Guzzetta, Andrea

AU - Vanhatalo, Sampsa

PY - 2019/3/18

Y1 - 2019/3/18

N2 - Abstract Aim General movement assessment requires substantial expertise for accurate visual interpretation. Our aim was to evaluate an automated pose estimation method, using conventional video records, to see if it could capture infant movements using objective biomarkers. Methods We selected archived videos from 21 infants aged 8-17 weeks who had taken part in studies at the IRCCS Stella Maris Foundation (Italy), from 2011-2017. Of these, 14 presented with typical low-risk movements, while seven presented with atypical movements and were later diagnosed with cerebral palsy. Skeleton videos were produced using a computational pose estimation model adapted for infants and these were blindly assessed to see whether they contained the information needed for classification by human experts. Movements of skeletal key points were analysed using kinematic metrics to provide a biomarker to distinguish between groups. Results The visual assessments of the skeleton videos were very accurate, with Cohen's K of 0.90 when compared with the classification of conventional videos. Quantitative analysis showed that arm movements were more variable in infants with typical movements. Conclusion It was possible to extract automated estimation of movement patterns from conventional video records and convert them to skeleton footage. This could allow quantitative analysis of existing footage. This article is protected by copyright. All rights reserved.

AB - Abstract Aim General movement assessment requires substantial expertise for accurate visual interpretation. Our aim was to evaluate an automated pose estimation method, using conventional video records, to see if it could capture infant movements using objective biomarkers. Methods We selected archived videos from 21 infants aged 8-17 weeks who had taken part in studies at the IRCCS Stella Maris Foundation (Italy), from 2011-2017. Of these, 14 presented with typical low-risk movements, while seven presented with atypical movements and were later diagnosed with cerebral palsy. Skeleton videos were produced using a computational pose estimation model adapted for infants and these were blindly assessed to see whether they contained the information needed for classification by human experts. Movements of skeletal key points were analysed using kinematic metrics to provide a biomarker to distinguish between groups. Results The visual assessments of the skeleton videos were very accurate, with Cohen's K of 0.90 when compared with the classification of conventional videos. Quantitative analysis showed that arm movements were more variable in infants with typical movements. Conclusion It was possible to extract automated estimation of movement patterns from conventional video records and convert them to skeleton footage. This could allow quantitative analysis of existing footage. This article is protected by copyright. All rights reserved.

KW - automatic pose estimation

KW - cerebral palsy

KW - general movements

KW - infants

KW - spontaneous motor activity

KW - 3123 Gynaecology and paediatrics

U2 - 10.1111/apa.14781

DO - 10.1111/apa.14781

M3 - Article

VL - 0

JO - Acta Paediatrica

JF - Acta Paediatrica

SN - 0803-5253

IS - ja

ER -