MUMBAI, India, Feb. 6 -- Intellectual Property India has published a patent application (202641007107 A) filed by Dayananda Sagar University; and Dayananda Sagar Academy Of Technology And Management, Bengaluru, Karnataka, on Jan. 24, for 'multi-stream gradient boosting lightgbm framework for alzheimer's disease diagnosis using mri and pet datasets.'

Inventor(s) include Prof. Pooja Shree H R; Dr. G Manjula; Lavanya K; Prof. Yashaswini H C; Prof. Gaurav Kumar; Prof. Kavyashree I Pattan; Prof. Bharath M B; and Prof. Mala B A.

The application for the patent was published on Feb. 6, under issue no. 06/2026.

According to the abstract released by the Intellectual Property India: "Alzheimer's disease is a progressive neurodegenerative disorder that requires early and reliable detection to enable timely clinical intervention. Traditional diagnostic methods often fail to identify subtle early-stage changes, creating the need for automated imaging-based solutions. Prior research indicates that MRI captures structural atrophy while PET identifies early metabolic abnormalities, and combining both modalities significantly enhances diagnostic accuracy. Building on these findings, this project aims to develop an automated system that analyzes MRI and PET scans from the ADNI-3 dataset to detect Alzheimer's disease with improved precision. The objective is to investigate multimodal feature extraction, test deep-learning-based CNN embeddings, and evaluate LIGHTGBM classification performance after fusing MRI and PET features. The methodology includes DICOM-TO NIFTI conversion, preprocessing through intensity normalization and resizing, and extraction of high-dimensional embeddings using convolutional neural networks. These fused features are then classified to determine the presence of Alzheimer's disease. A FASTAPI backend and web-based frontend are incorporated to enable user-friendly scan uploads and real-time predictions. Overall, the project seeks to assess the effectiveness of multimodal imaging and machine learning in supporting early, scalable, and efficient Alzheimer's detection."

Disclaimer: Curated by HT Syndication.