MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202641109847 A) filed by Vallurupalli Nageswara Rao Vignana Jyothi Institute Of Engineering And Technology on September 13, 2026, for An Ai-Driven Medical Image Description System Using Deep Learning And Intelligent Optimization Techniques.
Inventor includes Mrs. C. Divya.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: ABSTRACT [0021] An optimized medical image description system is disclosed that automatically produces coherent and contextually accurate captions for diagnostic images including X-rays, MRI scans, and CT scans. A pre-trained VGG16 network extracts high-level visual features from input images, which are then processed by a Long Short-Term Memory network to generate sequential word-by-word descriptions. Hyperparameters such as learning rate, number of LSTM units, and batch size are jointly refined through a hybrid procedure that merges Adaptive Coati Optimization with a Genetic Algorithm. The resulting model reduces overfitting, improves caption accuracy from approximately 0.87 to 0.94, and yields fluent, medically meaningful text. A web-based interface allows image upload and real-time display of the generated description, supporting faster and more consistent radiological interpretation in clinical settings.
Disclaimer: Curated by HT Syndication.
