Studies show telehealth methods for autism diagnosis are 80-91% accurate compared to in-person evaluations, with sensitivity 75-100% and specificity 68.75-100%13.
A December 2025 study on AI-assisted CanvasDx device produced no false positives or negatives in 52% of 80 children tested, accelerating diagnoses when used by experienced clinicians2.
Virtual assessments gained traction during COVID-19, offering benefits like increased accessibility, reduced waitlists, and better view of daily child behavior; clinicians report high confidence and plan continued use1.
Scoping review confirms telehealth matches in-person accuracy, e.g., 86-91% agreement on diagnoses and high sensitivity/specificity rates across multiple studies3.
New AI models combining eye-tracking with clinician assessments enable early diagnosis as young as 14 months in primary care settings4.
Sources:
1. https://www.abtaba.com/blog/virtual-autism-evaluation
2. https://www.news-medical.net/news/20251218/AI-assisted-evaluations-can-speed-autism-diagnosis.aspx
3. https://pmc.ncbi.nlm.nih.gov/articles/PMC8830614/
4. https://wsbt.com/news/local/new-technology-advances-in-early-autism-diagnosis-in-children-eye-tracking-technology-south-bend-elkhart-saint-joseph-county-primary-care-partners-neurodivergent-screening