One Home Video, Several Readings: How Firm Is an AI's Autism Screening Result?
AI tools that screen young children for autism give one verdict for each recording, and a parent cannot tell whether that verdict is firm. This note proposes that the screening program rebuild several versions of the same recording by moving its own processing settings within their permitted ranges, run the company's AI model on each version unchanged, and read the spread of the answers. The computer then decides a state from the spread, without waiting for a person: firm, ask again for a second recording, or send to a clinician. Parents and clinicians receive the result together with its firmness. The note describes what CogniAble, Cognoa and EarliTec have built, and where this idea fits beside them. Patent pending.
Authors
- Prakash Vaithyanathan (ORCID: https://orcid.org/0000-0001-8230-4830)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-10-31
- DOI
- https://doi.org/10.5281/zenodo.23846677
- Primary Topic
- Autism Spectrum Disorder Research
- Type
- article
- Field-Weighted Citation Impact
- 0.00