- The FDA has cleared Neurora, a biometric analysis tool that uses video and audio from standard assessments to help diagnose autism in children ages 18 months to 12 years.
- Research shows the tool correctly identifies autism in approximately 78% of cases, with higher accuracy when distinguishing autistic from neurotypical children.
- Autism now affects 1 in 31 U.S. children, with median diagnosis age near 4 years—delaying potential intervention.
- Earlier eye-tracking technology, EarliPoint, showed 78% sensitivity and 85% specificity in children ages 16-30 months.
- Critics and advocates emphasize that diagnostic tools, while valuable, should not overshadow foundational health priorities like nutrition, sleep and reducing environmental toxin exposures.
Federal regulators have cleared a new software tool this month designed to help clinicians identify autism in children. The Food and Drug Administration granted clearance for Neurora, developed by Argus Cognitive Inc., for children ages 18 months to 12 years. The device analyzes video and audio recordings from standard behavioral assessments, measuring facial expressions, voice patterns, eye gaze and more than 100 biometric markers. The results compare a child's behavior to patterns observed in children with and without autism.
The clearance represents the latest entry in a growing field of technology-assisted autism diagnostics. Earlier tools include EarliPoint, which uses eye-tracking technology to analyze visual behavior in children as young as 16 months, and Canvas Dx, which uses algorithmic analysis of parent-reported information and video submissions.
The science behind the screening
Researchers enrolled 546 participants across four sites in the United States and Qatar for the Neurora study. Of those, 458 met quality standards for analysis. The machine learning tool demonstrated 77.8% sensitivity and specificity in distinguishing between autistic and non-autistic children. When comparing autistic children to neurotypical participants only, accuracy rose to 82%.
The study noted performance differences by sex: males showed higher sensitivity (79%), while females showed higher specificity (84%). The tool struggled with differential diagnosis among children with other neurodevelopmental conditions that share autism-like features.
EarliPoint research, led by Dr. Warren Jones at Emory University, tested 475 children ages 16 to 30 months and found 78% accuracy for children with confirmed autism and 85% for those without.
Behind the numbers: What accuracy really means
The Centers for Disease Control and Prevention now estimates autism affects 1 in 31 U.S. children, with boys nearly four times more likely to be diagnosed. The median age of diagnosis is just under 4 years, despite symptoms often appearing before age 2.
For families, a 78% accuracy rate means roughly one in five children may receive incorrect results. The FDA classification requires these tools to inform clinical decision-making rather than provide standalone diagnoses. Physicians must still exercise professional judgment.
A broader perspective on child health
The push for earlier autism diagnosis comes as approximately one in six U.S. children ages 3-17 live with developmental disorders. Autism rates have risen from 1 in 150 children in 2000 to current estimates of 1 in 31.
While diagnostic tools evolve, many natural health advocates and families point to foundational factors often overlooked in clinical settings. Proper nutrition, adequate sleep and reducing exposure to environmental chemicals—including pesticides, phthalates and heavy metals—remain critical for childhood development. Research has linked prenatal and early childhood toxin exposure to neurodevelopmental challenges, yet mainstream diagnostic conversations rarely address these contributors.
The diagnostic label itself, advocates argue, should serve as a starting point rather than an endpoint. Families benefit most when assessment results translate into concrete plans for care and support, including addressing underlying physiological and environmental factors that may influence behavior and development.
A new chapter in diagnostic technology
Neurora's development received support from a National Institutes of Health Small Business Innovation Research grant. The tool does not require clinicians to change existing assessment protocols or purchase proprietary hardware, potentially lowering barriers to adoption.
However, researchers acknowledged that "known difficulties with differential diagnosis in non-ASD neurodevelopmental conditions with autism-like features remain a limitation requiring further development."
Looking forward
The availability of objective, non-invasive diagnostic tools represents progress for families currently facing average wait times of years for autism evaluations. Faster identification can open doors to early interventions and support services.
Yet technology alone cannot address the full picture of childhood development. As diagnostic precision improves, families and clinicians alike may benefit from remembering that labels describe behaviors, not children, and that optimal health outcomes require attention to the whole child—including what they eat, how they sleep, and the environment in which they grow.
Sources for this article include:
Please contact us for more information.