MUMBAI, India, July 30 -- Intellectual Property India has published a patent application (202641088768 A) filed by Dr. Sridhar C S; Mrs. Roopa G, Hod, Aiet; Ms. Meghana, Ap, Aiet; Dr. Thritha Prasada H P, Hod, Vtu Muddenahalli; Ms. Pratiksha, Ap, Aiet; Mr. Sunil, Ap; Mrs. S P Preethi, Ap, Aiet; and Ms. Harshitha G J, Ap, Aiet on July 21, 2026, for Large Vector Based Semantic Search Engine Using Retrieval Augmented Generation For Medical Literature.
Inventors include Dr. Sridhar C S; Mrs. Roopa G, Hod, Aiet; Ms. Meghana, Ap, Aiet; Dr. Thritha Prasada H P, Hod, Vtu Muddenahalli; Ms. Pratiksha, Ap, Aiet; Mr. Sunil, Ap, Aiet; Mrs. S P Preethi, Ap, Aiet; and Ms. Harshitha G J, Ap, Aiet.
The application for the patent was published on July 24, 2026, under issue no. 30/2026.
Abstract: This system integrates Natural Language Processing (NLP), transformer-based embedding models, vector databases, semantic search techniques, and Large Language Models (LLMs) to overcome the limitations of conventional keyword-based medical search systems. Medical documents are collected from trusted sources, including research journals, clinical practice guidelines, systematic reviews, medical textbooks, electronic health records, and clinical trial repositories. The collected documents are preprocessed through text normalization, tokenization, and medical entity extraction before being converted into high-dimensional semantic vector embeddings and stored in a scalable vector database. The retrieved information is supplied to a Retrieval-Augmented Generation (RAG) module, where a Large Language Model generates concise, evidence-based, and contextually accurate responses supported by the retrieved literature. The system further incorporates explainable AI by presenting supporting references, document citations, and confidence scores, enabling users to verify the generated information.
Disclaimer: Curated by HT Syndication.
