Toward Novel Tools for Autism Identification: Fusing Computational and Clinical Expertise.
Laura L Corona, Liliana Wagner, Joshua Wade and 6 others
PMID 33417138WHAT IT FOUND
Researchers built a brief play-based observation tool for toddlers using machine learning to find the most predictive behaviors.
The tool focuses on seven areas like eye contact and vocalization. It is not yet validated for clinical diagnosis but shows promise for identifying symptoms quickly.
Key findings
01The study developed a novel, brief assessment tool based on seven behavioral constructs identified as most predictive of ASD in a database of 737 toddlers.
02The tool's design aims for administration in under 20 minutes using low-cost, widely available materials to facilitate use in community practice settings.
03The predictive behaviors selected by the model include socially directed speech, frequent eye contact, unusual vocalizations, repetitive play, repetitive body movements, combined gestures and eye contact, and unusual sensory reactions.
STILL TO COME
How it was doneWhat they foundWhat it means for SLPs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The study describes the development of a tool but does not report on its accuracy or validity in a new sample. The underlying data comes from a specialized clinical database and may not represent the general population. The dataset lacked information on race, ethnicity, and medical complexity, limiting understanding of how the tool might perform across diverse groups. The new tool is currently in early stages of investigation across various settings (apps, telehealth) with no published outcome data yet.
Declared interests
Funded by NIH/NIMH, NIH, NICHD, and the Center for Clinical and Translational Sciences at UTHealth Houston.
The easy way to misread this
Do not use this tool in clinical practice to diagnose or screen patients. It is a description of a prototype's development, not a validated instrument with published accuracy rates for the new procedure itself.