MUMBAI, India, May 29 -- Intellectual Property India has published a patent application (202641063428 A) filed by M. Kumarasamy College Of Engineering, Karur, Tamil Nadu, on May 19, for 'adverse drug reaction prediction using machine learning on patient and drug data.'
Inventor(s) include Dr. M. Murugesan; Aswin Senthilkumar; Arul Murugan S B; and Bhavanithi K.
The application for the patent was published on May 29, under issue no. 22/2026.
According to the abstract released by the Intellectual Property India: "The present invention discloses a computer-implemented system for the early detection and probability estimation of adverse drug reactions through hybrid artificial intelligence techniques. The system integrates both structured data (including patient demographics, drug names, and reported effects) and unstructured clinical text (including patient narratives and medical notes) to identify potential ADRs before they become clinically evident. The invention employs Natural Language Processing (NLP) for text preprocessing, including tokenization, lemmatization, and Named Entity Recognition (NER) to extract biomedical entities such as drugs and symptoms. For structured inputs, features are encoded and normalized to ensure uniform representation. These multimodal features are fused and processed by an XGBoost-based machine learning classifier, which computes an ADR probability score and a binary risk classification The system produces interpretable outputs, highlighting the detected drug and symptom entities alongside their corresponding ADR probabilities. A web-based interface allows real-time prediction via text input or batch processing through CSV upload. The invention operates as a software-only, scalable, and hardware-independent solution that supports pharmacovigilance, clinical research, and healthcare analytics by improving the accuracy, speed, and interpretability of ADR detection."
Disclaimer: Curated by HT Syndication.
