MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202641094592 A) filed by Sai Krishna Prasad Munnaluri; and Abdul Nazeer K. A on August 04, 2026, for An Explainable Agentic Artificial Intelligence Framework For Interpreting Autism Screening Data.
Inventors include Sai Krishna Prasad Munnaluri; and Abdul Nazeer K. A.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: The present invention relates to a computer-implemented explainable agentic artificial intelligence framework for interpreting structured autism screening data. The framework comprises an input interface configured to receive a structured developmental screening record, a data interpreter agent configured to validate the received responses, map the validated responses to one or more developmental domains, calculate domain-level aggregates, and construct an ordered evidence context, a plurality of specialized generation agents configured to independently generate caregiver-readable report sections from the ordered evidence context, an orchestration mechanism configured to aggregate the generated report sections, and a validation agent configured to independently evaluate each generated report section for faithfulness to the ordered evidence context, consistency with operational score and label logic, and compliance with predefined non-diagnostic communication requirements. The validation agent further identifies unsupported content, inconsistencies or policy violations and selectively repairs, revalidates, flags or withholds report sections prior to generation of a validated caregiver-readable report. The disclosed framework provides a modular, explainable and traceable processing architecture that improves reliability and transparency in communication of structured autism screening information while maintaining consistency between generated explanatory outputs and the underlying screening responses. FIG 01.
Disclaimer: Curated by HT Syndication.
