Research Projects

Objective: A. Development of an online or locally deployable AI system for detection of Intracranial Bleeds B. Validation of the developed AI model and system in a clinical setting C. Development of a methodology to enable generalization of the AI model for multiple hospitals D. Extending the developed methods onto other different conditions

Summary: AI, despite having tremendous potential, has seen relatively low implementation in medical practice across India. Standard procedure for clinical triage emergencies such as CT or X-ray requires a radiologist which are either unavailable in remote areas or in high demand where available leading to delay in reports for possibly fatal conditions while posing significant stress on the radiologists. Recently in a previous project, we have successfully built and implemented deep learning (DL) models for early detection of intracranial bleeds having 95% accuracy. In this proposal, we intend to deploy and validate DL models in hospitals while extending our methods to triage of ureteric stones and gastrointestinal perforation.