ASC-emotion: A privacy-aware dataset for analysing emotional dysregulation and engagement in children with Autism()

09/08/2026

AI Summary

Curated privacy-preserved Autism meltdown dataset enables early detection, standardised benchmarking, reproducibility and translational health outcomes for challenging behaviours in children with ASC.
Privacy preserving measures integrated into the machine learning pipeline mitigate ethical concerns using video data from a vulnerable population.
Empirical validation with BiLSTM, EdgeConv GNN and PointCNN+LSTM models achieved 96% accuracy detecting learning related affective states and physiological arousal.

AI Summary

The dataset was validated using three machine learning architectures which gave an accuracy of 96% in detecting the affective states related to learning in children with Autism.

Development of an Autism meltdown dataset exemplifies methodological rigor, advances an urgent clinical challenge through early detection and intervention, and enables broad impact by promoting reproducibility, benchmarking and translational health outcomes. 2.

High levels of accuracy are demonstrated via empirical validation of the dataset through three machine learning models (BiLSTM, Graphical Neural Network - EdgeConv, PointCNN+LSTM) for detecting affective states related to learning and physiological arousal in children with Autism.

Basic summary

MethodsX. 2026 Jul 27;17:104073. doi: 10.1016/j.mex.2026.104073. eCollection 2026 Dec.

ABSTRACT

Predicting emotional dysregulation events in children with Autism is essential for timely mitigation of triggering events and prevention of further escalation of the situation. However, there is a scarcity of accessible and standarised datasets for use in AI-based research associated with challenging behaviours in children with ASC. To address this gap, we have curated a novel privacy-preserved dataset as part of an Erasmus+ funded project (AI-TOP-2020-1-UK01-KA201-079,167). The dataset was validated using three machine learning architectures which gave an accuracy of 96% in detecting the affective states related to learning in children with Autism. The key contributions of this study are: 1. Development of an Autism meltdown dataset exemplifies methodological rigor, advances an urgent clinical challenge through early detection and intervention, and enables broad impact by promoting reproducibility, benchmarking and translational health outcomes. 2. Implementation of privacy preserving measures addresses ethical concerns regarding the use of video data with this vulnerable population as part of the machine learning pipeline. 3. High levels of accuracy are demonstrated via empirical validation of the dataset through three machine learning models (BiLSTM, Graphical Neural Network – EdgeConv, PointCNN+LSTM) for detecting affective states related to learning and physiological arousal in children with Autism.

PMID:42569523 | PMC:PMC13450243 | DOI:10.1016/j.mex.2026.104073

Document this CPD