MUMBAI, India, June 22 -- Intellectual Property India has published a patent application (202641070274 A) filed by Cvr College Of Engineering on June 05, 2026, for An Automated Visual-Based Mental Health Diagnosis System Using Hierarchical Deep Learning And Reliability-Aware Feature Integration.
Inventor includes Hari Shankar Punna.
The application for the patent was published on June 12, 2026, under issue no. 24/2026.
Abstract: The current invention concerns a visual-based system that leverages deep learning algorithms to perform a mental health diagnosis automatically and reliably. Specifically, the proposed method is called NeuroVision-MH and comprises hierarchical feature extraction, feature integration with adaptivity, and reliability-aware machine learning to perform diagnosis on images related to facial expressions and neural scans with regards to such mental conditions as depressive, anxiety disorders, mild cognitive impairment, and Alzheimer's illness. To start with, input images are preprocessed through normalization, improvement, and data augmentation procedures, and subsequently processed via deep convolutional learning networks that provide multiple levels of feature representations. Adaptively weighted features ensure that only meaningful facial features contribute to the decision process, while reliability assessment helps measure prediction certainty and uncertainty. This system is capable of performing reliable, accurate, robust, and interpretable automatic diagnostic procedures that minimize uncertainties of predictions and are better than CNN and pre-trained deep learning algorithms.
Disclaimer: Curated by HT Syndication.
