[Paper Review] A Review of and Roadmap for Data Science and Machine Learning for the Neuropsychiatric Phenotype of Autism

Peter Washington, Dennis P. Wall|arXiv (Cornell University)|Mar 7, 2023

Autism Spectrum Disorder Research4 citations

TL;DR

This paper reviews data science and machine learning applications for digital phenotyping and clinical management of autism spectrum disorder (ASD), focusing on behavioral quantification, diagnostic classification, and therapeutic interventions. It outlines a roadmap for translational deployment, identifying key challenges and opportunities in leveraging ML for longitudinal monitoring and personalized care in heterogeneous neuropsychiatric phenotypes.

ABSTRACT

Autism Spectrum Disorder (autism) is a neurodevelopmental delay which affects at least 1 in 44 children. Like many neurological disorder phenotypes, the diagnostic features are observable, can be tracked over time, and can be managed or even eliminated through proper therapy and treatments. Yet, there are major bottlenecks in the diagnostic, therapeutic, and longitudinal tracking pipelines for autism and related delays, creating an opportunity for novel data science solutions to augment and transform existing workflows and provide access to services for more affected families. Several prior efforts conducted by a multitude of research labs have spawned great progress towards improved digital diagnostics and digital therapies for children with autism. We review the literature of digital health methods for autism behavior quantification using data science. We describe both case-control studies and classification systems for digital phenotyping. We then discuss digital diagnostics and therapeutics which integrate machine learning models of autism-related behaviors, including the factors which must be addressed for translational use. Finally, we describe ongoing challenges and potent opportunities for the field of autism data science. Given the heterogeneous nature of autism and the complexities of the relevant behaviors, this review contains insights which are relevant to neurological behavior analysis and digital psychiatry more broadly.

Motivation & Objective

• To synthesize existing research on data science and machine learning for digital phenotyping of autism spectrum disorder (ASD).
• To identify critical bottlenecks in current diagnostic, therapeutic, and longitudinal tracking pipelines for ASD.
• To evaluate the state of machine learning models in classifying autism-related behaviors and their potential for clinical integration.
• To propose a roadmap for translating data science innovations into real-world clinical workflows for ASD.
• To highlight cross-cutting challenges and opportunities relevant to digital psychiatry and neurological behavior analysis.

Proposed method

• Systematic review of peer-reviewed literature on data science and machine learning applications in autism research.
• Categorization of studies into case-control designs and classification systems for digital phenotyping using behavioral data.
• Analysis of machine learning models trained on multimodal data (e.g., speech, movement, eye-tracking) to detect autism-related phenotypes.
• Evaluation of technical, ethical, and clinical factors affecting the translational potential of ML models in ASD diagnostics and therapeutics.
• Synthesis of challenges related to data heterogeneity, model interpretability, and longitudinal tracking in real-world settings.
• Development of a structured roadmap for future research and clinical deployment, emphasizing scalability and equity in access.

Experimental results

Research questions

• RQ1 What are the current state-of-the-art data science and machine learning methods for quantifying autism-related behaviors?
• RQ2 How do existing digital phenotyping systems perform in classifying autism spectrum disorder across diverse populations?
• RQ3 What technical, ethical, and clinical barriers impede the real-world deployment of ML-based diagnostics and therapeutics for ASD?
• RQ4 How can data science approaches be optimized for longitudinal monitoring of neuropsychiatric phenotypes in autism?
• RQ5 What are the key opportunities for expanding access to autism services through scalable data-driven solutions?

Key findings

• A growing body of research demonstrates the feasibility of using machine learning to classify autism-related behaviors from digital phenotyping data, including speech, movement, and social interaction patterns.
• Many existing models show promising classification performance in controlled settings, though generalizability across diverse populations remains limited.
• Key challenges include data heterogeneity, lack of standardized metrics, and insufficient attention to model interpretability and clinical utility.
• Longitudinal tracking of behavioral phenotypes using data science tools is underexplored but holds significant potential for early intervention and personalized therapy.
• There is a critical need for standardized benchmarks, ethical frameworks, and collaborative data-sharing initiatives to accelerate clinical translation.
• The insights from this review are broadly applicable to digital psychiatry and neurological behavior analysis beyond autism.