Replication Data for: Markerless motion capture to assess upper extremity movements in individuals with dyskinetic cerebral palsy: an accuracy and validity study

The dataset contains 2D coordinates in pixels of key points extracted from videos of 30 children, adolescent and young adults with dyskinetic cerebral palsy (DCP) and 21 typically developing (TD) peers (age 5-24). The performed task was a reaching-sideways task (from starting point on the upper leg to the point of task achievment (PTA) sidewards at shoulder height) including eight repetitions. The key points were extracted from videos using open-source toolbox code DeepLabCut version 2.3 [1, 2]. The collected key points were: ankle, knee, hip, shoulders, elbow, wrist, forehead and chin, metacarpophalangeal III joint (MCP3) and the reaching target.

In addition, the joint angles collected with three-dimensional movement analysis (3DMA) at PTA are added - used for validation purposes in the related article.

The videos and 3DMA data were collected during a previous study [3].

Data belong to the project 'Automated VIdeo-based assessment of DYSkinesia in cerebral palsy using markerless pose estimation and machine learning (AVI-DYS)' and 'Instrumented dystonia and choreoathetosis assessment protocol (IDCA) of upper limb movements in cerebral palsy' .

References:

1. Mathis, A., et al., DeepLabCut: markerless pose estimation of user-defined body parts with deep learning. Nat Neurosci, 2018. 21(9): p. 1281-1289.

2. github.com/DeepLabCut/DeepLabCut

3. Vanmechelen I, et al., Psychometric properties of upper limb kinematics during functional tasks in children and adolescents with dyskinetic cerebral palsy. PLoS One 2022;17.

Jaar van publicatie:2023

Toegankelijkheid:open

Uitgever:KU Leuven RDR

Licentie:CC-BY-4.0

Formaat:ipynb, xlsx

Trefwoorden: Adolescent, Cerebral palsy, dyskinetic, Child, Deep Learning, Movement, Upper Extremity, Young Adult