PCPP: A MATLAB application for abnormal infant movement detection from video

Ho, ESL, McCay, KD, Marcroft, C and Embleton, ND (2022) PCPP: A MATLAB application for abnormal infant movement detection from video. Software Impacts, 14. p. 100412.

Preview

Published Version Available under License Creative Commons Attribution. Download (554kB) | Preview

Abstract

PCPP is an application developed in MATLAB, for the detection of abnormal infant movements associated with cerebral palsy. This system uses 2D skeletal data extracted from videos, and consists of a full pipeline providing data pre-processing, data normalization, feature extraction and classification. Evaluation metrics, such as accuracy, sensitivity, specificity, F1 score and Matthews Correlation Coefficient (MCC), are computed to facilitate full assessment of performance and allow for comparison with other methods from the literature. These evaluations are conducted on the MINI-RGBD and RVI-38 datasets using the code and data provided.