IIIT-H Helps Build AI That Bridges Info Gap Between Cancer Patients And Doctors
Hyderabad: Cancer patients often leave hospitals with questions that arise only after they return home; but a new multilingual artificial intelligence (AI) platform developed by Christian medical college (CMC), Vellore and international institute of information technology Hyderabad (IIIT Hyderabad) aims to bridge this gap by answering routine questions from patients in their own language while helping doctors understand developing symptoms between appointments.
The platform, called BandhuCare, allows patients to speak or write naturally about symptoms such as pain, difficulty swallowing or dry mouth. Rather than simply recording responses, AI transforms everyday conversations into structured clinical information that doctors can review before the next exam.
Prof. from IIIT Hyderabad’s language technologies research centre. “BandhuCare is not designed to diagnose diseases or prescribe treatment, but to answer routine questions, explain symptoms commonly experienced by patients, and refer patients back to the healthcare team when clinical intervention is needed,” said Dipti Misra Sharma.
Unlike general-purpose AI chatbots, the system only uses hospital-approved medical information. D., scientist and leader of the quantitative imaging research and artificial intelligence laboratory at CMC. “The answer comes from clinically validated guidelines, not generic online advice. Before any answer reaches the patient, another layer of AI checks whether the answer actually comes from the validated knowledge base,” said Hannah Mary Thomas.
The platform currently supports eight Indian languages via text and voice using IIIT Hyderabad’s language technologies. It also includes a symptom diary that allows patients to record their daily experiences; This helps doctors get a clearer picture of what happened between visits rather than relying solely on memory.
D., professor and chief of radiation oncology at CMC. Balu Krishna said that patients often leave the examination with unanswered questions, and language barriers further limit access to reliable information. “We want the technology to adapt to patients rather than the other way around,” Dr Thomas added.
