Canvas Dx

Canvas Dx is a prescription diagnostic aid that helps healthcare providers diagnose Autism Spectrum Disorder (ASD) in children ages 18-72 months at risk for developmental delay. It combines a caregiver mobile app, video analysis, and healthcare provider questionnaire inputs into a machine learning algorithm that outputs whether the patient is positive, negative, or indeterminate for autism. The device is designed to support clinical decision-making, not replace professional diagnosis.

Canvas Dx is identical to the predicate device (Cognoa ASD Diagnosis Aid, DEN200069) in terms of regulatory class, product code, intended use, data collection methods, outputs, and software standards (ISO 62304 compliance). The primary difference is inclusion of a Predetermined Change Control Plan (PCCP) that permits future updates to machine learning model performance thresholds and decision algorithms without requiring new 510(k) submissions, provided changes remain within the established bounds.

Testing was performed in accordance with IEC 62304:2006 (medical device software lifecycle processes) to demonstrate safety based on current industry standards. Software complies with ISO 62304.

Canvas Dx is substantially equivalent because it is functionally identical to the predicate device (DEN200069)—same manufacturer, same regulatory class and product code, same indications for use, same data inputs and outputs, and same software compliance standards. The device was simply renamed from Cognoa ASD Diagnosis Aid to Canvas Dx shortly after the original De Novo authorization. The clinical performance data from the predicate's validation study (sensitivity 98.44%, specificity 78.87%, PPV 80.77%, NPV 98.25%) remains valid for the subject device since no technological changes have occurred in the core device functionality or intended use.