~ In an interaction with ETHealthworld, Vasudevan Sundarababu, Senior Vice President and Chief Technology officer (CTO) at CSS Corp, shares insight on developments in conversational artificial intelligence and how it is redefining the healthcare scenario ~
AI does improve and cut down on the human error in terms of diagnosis and relapse however the final decision is taken by the doctors. In an interaction with ETHealthworld, Vasudevan Sundarababu, Senior Vice President and Chief Technology officer (CTO) at CSS Corp, shares insight on developments in conversational artificial intelligence and how it is redefining the healthcare scenario.
Your views on conversational AI adding value to healthcare
Conversational AI is revolutionizing patient interactions primarily in booking appointments, obtaining drug information and receiving healthcare reminders like flu vaccination. Conversational AI and chatbots are the last mile of digitization, before that comes contextualization. Contextualization is the new personalization. So how do I understand the context or need of a patient? That happens with the digitization and creation of a robust knowledge base, which voice-based channels can tap into.
The accuracy of the technology revolves around deep understanding of natural language processing and replying in a contextual manner. This technology is rapidly evolving and I think in the next 2-3 years, the usage would be more prevalent as it would reduce waiting period in the hospitals, help in explanation of drugs in prescriptions and so on. So, the usage and adoption will be very high going forward.
Having said that, in the near future, conversational AI is not going to prescribe a medicine to patients because prescription of medication involves consideration of multiple factors beyond just the codified data. Conversational AI isn’t intelligent enough to replace doctors and we are still far away from that point.
How ML and AI improve diagnosis?
AI is currently used as an augmentation tool for doctors, basis which they can take informed decisions. There are many use cases where it has been effective. Importantly, AI helps us to reduce false negative results. For instance, if there is a tumor which can be cancerous, and the results are false negative, AI helps to detect that and report it. It helps to flag scenarios which escape human eye or human diagnosis. AI can be used to analyze various images like CT Scans and X-rays. So, AI does improve and cut down on human error in terms of diagnosis.
Another place where machine learning and AI helps us is in relapse. The system analyses and tracks cases of a possible relapse occurring which may not be normally identified by physicians. These cases are spotted and post-surgical care can be also suggested through ML and AI technologies. But reiterating that the final decision is still taken by doctors, so human intervention is still required.
What are the limitation and challenges of AI in the current Indian healthcare setting?
India is steadily adopting AI, but the adoption in healthcare as compared to the West is still relatively slower as it is mainly to do with digitization. AI needs data to work. The first step should be digitizing all patient health records from x-rays to prescription records to truly harness AI capabilities. Digitization and AI would also tremendously help in other healthcare areas like drug development, clinical trials and analysis of surgical procedures.