MUMBAI, India, Sept. 7 -- Intellectual Property India has published a patent application (202641104879 A) filed by Mrs. Sabiha Nuzhat; and Dr. Shubhangi D. C. on August 31, 2026, for An Explainable Ai-Based Framework Early Detection Of Autism Spectrum Disorder Using Machine Learning Techniques.
Inventors include Mrs. Sabiha Nuzhat; and Dr. Shubhangi D. C..
The application for the patent was published on September 04, 2026, under issue no. 36/2026.
Abstract: The present invention provides an explainable artificial intelligence-based system and method for early detection of Autism Spectrum Disorder using multimodal machine learning. The system comprises a data acquisition interface for receiving heterogeneous subject-specific information, an adaptive preprocessing engine for performing modality-dependent preprocessing and determining data-quality characteristics, a feature engineering engine, a feature relevance determination engine, a machine learning inference engine, an explainability engine, a reliability evaluation engine and an assessment output interface. The machine learning inference engine generates an autism-associated assessment value from selected features and may generate modality-specific outputs that are fused according to dynamically determined modality weights. The explainability engine determines feature contributions responsible for the generated assessment. The reliability evaluation engine determines a reliability indicator using prediction confidence, input-data quality, distribution conformity and cross-modal agreement. The system outputs the assessment value together with the reliability indicator and human-interpretable explanatory information and may flag uncertain or inconsistent assessments for further professional evaluation. Accompanied Drawing [FIGS. 1-2]
Disclaimer: Curated by HT Syndication.
