Cognoa ASD Diagnosis Aid

The Cognoa ASD Diagnosis Aid is a software-based medical device designed to facilitate earlier identification of Autism Spectrum Disorder in children aged 18-72 months by integrating caregiver reports, clinician observations, and video-based behavioral analysis through a machine-learning algorithm. This system, trained on diverse pediatric data and demonstrating notable sensitivity and specificity, aims to provide clinicians with an objective risk score to improve referral decisions and expedite access to critical early interventions.

The Cognoa ASD Diagnosis Aid is a sophisticated software tool developed to support clinicians in the evaluation of Autism Spectrum Disorder (ASD) for children aged 18 to 72 months. Its core methodology involves a multi-modal data integration approach: it collects information from caregiver-completed questionnaires, data entered by clinicians via a dedicated portal, and parent-recorded videos of the child's behavior. These videos are subsequently reviewed and annotated by trained analysts, providing structured observational data which, combined with the other inputs, forms the basis for the diagnostic assessment. The scientific foundation of the Cognoa aid lies in a machine-learning classifier, which processes the integrated data inputs to generate a quantifiable ASD risk score. This score categorizes results as Positive for ASD, Negative for ASD, or No Result, thereby assisting clinicians in making informed referral decisions for further diagnostic workup or intervention. Validated through pivotal studies involving a comprehensive dataset of children with and without autism diagnoses, the device has demonstrated approximately 83% sensitivity and 83% specificity. It is intended for use as an adjunctive tool within standard diagnostic workflows, aiming to facilitate earlier ASD identification and improve access to interventions, particularly in community healthcare settings.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. How does the 'No Result' output specifically guide clinicians, and what are the predominant factors leading to this outcome in the Cognoa ASD Diagnosis Aid?; What are the specific behavioral features or patterns extracted from the parent-recorded videos that most heavily influence the machine learning model's ASD risk assessment?; Beyond overall sensitivity and specificity, how does the diagnostic aid perform across diverse demographic subgroups, including different socioeconomic backgrounds or children with comorbid developmental conditions?; What is the training and inter-rater reliability protocol for the analysts reviewing parent-recorded videos, and how is this quality control maintained over time?; Are there comparative studies evaluating the real-world impact of implementing the Cognoa ASD Diagnosis Aid on metrics such as time-to-diagnosis and access to early intervention services versus standard care pathways?