MUMBAI, India, June 26 -- Intellectual Property India has published a patent application (202641072848 A) filed by Cvr College Of Engineering on June 12, 2026, for Transformer-Based Framework For Autism Detection Using Eeg Signals.
Inventors include Moghal Yaseen Pasha; Shaik Janbhasha; Moghal Irfan Pasha; and Nallametti Nagarjuna.
The application for the patent was published on June 19, 2026, under issue no. 25/2026.
Abstract: EEG signals are crucial for the early detection of neurological abnormalities associated with autism spectrum disorder. The existing machine learning and convolution methods for autism detection typically struggle to capture complex synchronisations of neural activity and long-range temporal dependencies from multi-channel EEG data. This invention introduces a Transformer-Based Framework for Autism Detection Using EEG Signals, which is able to model temporal and contextual dependencies within the sequences of EEG signals, and autonomously identify neural activity patterns associated with autism, which allows efficient contextual learning, better representation of features and detection of subtle neurological abnormalities. The framework is built using EEG data from OpenNeuro Autism EEG Dataset, which will be used for autism classification and neurological assessment. The proposed system extends autism detection accuracy, robustness, scalability, interpretability and real-time performance for implementation in hospital, wearable EEG platforms, and tele-environments for automated neurological assessment and early autism screening.
Disclaimer: Curated by HT Syndication.
