Automatic Detection of ADHD and ASD from Expressive Behaviour in RGBD Data

Abstract

Attention Deficit Hyperactivity Disorder (ADHD) and Autism Spectrum Disorder (ASD) are neurodevelopmental conditions which impact on a significant number of children and adults. Currently, the diagnosis of such disorders is done by experts who employ standard questionnaires and look for certain behavioural markers through manual observation. Such methods for their diagnosis are not only subjective, difficult to repeat, and costly but also extremely time consuming. In this work, we present a novel methodology to aid diagnostic predictions about the presence/absence of ADHD and ASD by automatic visual analysis of a persons behaviour. To do so, we conduct the questionnaires in a computer-mediated way while recording participants with modern RGBD (Colour+Depth) sensors. In contrast to previous automatic approaches which have focussed only detecting certain behavioural markers, our approach provides a fully automatic end-to-end system for directly predicting ADHD and ASD in adults. Using state of the art facial expression analysis based on Dynamic Deep Learning and 3D analysis of behaviour, we attain classification rates of 96% for Controls vs Condition (ADHD/ASD) group and 94% for Comorbid (ADHD+ASD) vs ASD only group. We show that our system is a potentially useful time saving contribution to the diagnostic field of ADHD and ASD.

A Survey on Computer Vision for Assistive Medical Diagnosis From Faces

“…As ADHD is characterized by symptoms such as hyperactivity, impulsivity, inattention, etc., Kinect has been suggested as a modality to assess behaviors of patients under specific tasks [140]. Eventually, Jaiswal et al [141] created the database KOOMA, containing video and Kinect recordings of 55 subjects (controls, patients with ADHD / autism) listening, reading stories and answering questions. Using facial expression analysis and 3D analysis of behavior, they were able to reach a classification accuracy of 96% for controls vs condition groups (ADHD/autism), these disorders sharing some common symptoms between each other's.…”

Classification of Attention Deficit Hyperactivity Disorder (ADHD) Considering Diagnosis and Treatment

“…The study [23] analysis of EEG data uses mutual information and the classifier achieves 85.7% accuracy to detect ADHD in children. The work [24] uses dynamic deep learning technique to segregate the ADHD and ASD patients separately in very less time duration with 96% of classification rate. The study [25] uses EEG signal to identify ADHD and achieved high accuracy of 87.5% in rest state.…”

Automatic Identification of High-Risk Autism Spectrum Disorder: A Feasibility Study Using Video and Audio Data Under the Still-Face Paradigm

“…For example, Jaiswal et al [13] designed a paradigm with adult subjects reading and listening to short stories, after which they proposed using computer vision cues derived from RGB-D data as features for detection of ASD and attention-deficit/hyperactivity disorder (ADHD). Liu et al [14] developed a machine learning method for identifying ASD for 4-to 11-year-old children through tracked eye-movement data, which was collected in an experimental scenario where children were asked to distinguish between two races based on facial images.…”

“…Guha et al [16] proposed a computational approach to reveal the facial expressions imitation details at 9-14 years of life for high-functioning autism (HFA) and TD children, where the reduced complexity in dynamic facial behaviors was found to arise primarily from the eye region for those HFA children. Although the existing researches [13], [14], [15], [16] focusing on an automatic diagnosis of ASD have achieved some progress, yet these studies were based on comparatively older subjects who belonged to groups of children, teenagers or adults. Some aforementioned experimental paradigms and methods are even not applicable to the babies before 24 months of age, because their language skills, behavioral abilities and IQs are still in development.…”