Machine Learning Analysis of fNIRS Data to Predict Therapy Engagement in Young Children

Presenter

Alex Dhima

Campus

UMass Amherst

Sponsor

Adam S. Grabell, Department of Psychological and Brain Sciences, UMass Amherst

Schedule

Session 5, 3:30 PM - 4:15 PM [Schedule by Time][Poster Grid for Time/Location]

Location

Poster Board A74, Campus Center Auditorium, Row 4 (A61-A80) [Poster Location Map]

Abstract

Artificial Intelligence (AI) is now being used in many sub disciplines of clinical mental health, from neuroimaging analysis to novel methods of disorder diagnoses. While a variety of studies and softwares have elicited beneficial and efficient results, there remains a paucity of AI applications to early childhood populations. A lack of comprehensive diagnostic criteria and accessible resources in early childhood has prevented approximately half of children who suffer from a mental health disorder from receiving needed treatment. There is hence enormous potential for AI to efficiently process multimodal data and provide concise analyses to address early childhood underdiagnosis and undertreatment. In combination with the emergence of fNIRS as a portable and non-invasive alternative to expensive neuroimaging techniques, AI could help provide personalized insights to clinicians on mental health diagnoses and treatment routing from a child's brain scan during a short therapeutic-based task. The current study explores the possibility of this precision-based model by first determining if neural activation patterns can demarcate engagement in a psychotherapy or non-therapy activity. Leveraging multiple machine learning algorithms, we aim to develop a classification model to detect whether a given child is engaged in a psychotherapy task from their in-vivo neural activity. The development of a predictive model from short brain scans may help clinicians reveal emerging forms of psychopathology and treatment responsiveness, advancing more tailored, neuroscience-informed care for young children and their families.

Keywords

artificial intelligence, neuroscience, emotion regulation, Functional Near-Infrared Spectroscopy, therapy

Research Area

Psychology and Behavioral Sciences

SIMILAR ABSTRACTS (BY KEYWORD)