MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202641112623 A) filed by Vallurupalli Nageswara Rao Vignana Jyothi Institute Of Engineering And Technology on September 19, 2026, for A Multimodal Deep Learning Framework Using Generative Adversarial Networks For Early Alzheimer’s Disease Detection And Diagnosis.
Inventor includes Mr. K. Kishan Babu.
The application for the patent was published on September 25, 2026, under issue no. 39/2026.
Abstract: ABSTRACT [0021] A multimodal fusion method based on a conditional generative adversarial network is presen ted for diagnosing Alzheimer's disease at an early stage. Structural T1-weighted MRI and functional 18F-FDG PET scans obtained from the Alzheimer's Disease Neuroimaging Initiative undergo initial co-registration and preprocessing, then passed through a dual-encoder generator that extracts modality-specific features and merges them by element-wise addition. A PatchGAN discriminator evaluates the realism of the fused output at multiple spatial scales. The resulting fused images retain both anatomical detail and metabolic information and are subsequently classified by lightweight convolutional neural networks, sorting them into cognitively normal, mild cognitive impairment, or Alzheimer's disease categories . Experiments on a balanced set of one hundred five subjects demonstrate that the fused representations yield higher classification accuracy than either modality used alone, thereby supporting more reliable early detection when complete imaging data are available.
Disclaimer: Curated by HT Syndication.
