Why health care will benefit the most from AI revolution

AI proliferation in medical sector depends on how its appetite for data is satiated


CIM HAS NOT attended medical college. It is not even human. But the AI-based engine is way more equipped to give a better diagnosis than you googling your symptoms and hoping to play doctor. Recently integrated into the clinics at Apollo Hospitals, CIM, or Clinical Intelligence Machine, analyses symptoms, determines causes and recommends the best course of action for a patient.

The engine is only part of the many AI-based initiatives implemented by Apollo across the country to “improve diagnosis, doctor productivity and patient satisfaction,” said Sangita Reddy, joint managing director of Apollo Hospitals. “We have moved a step forward for better patient experience. Technology is transforming the face of healthcare, making it more patient-centric and also reducing the burden on health care professionals.”

Startups and tech companies have been devising AI and ML tools that can assist in healthcare, both preventive and curative. “From the perspective of AI for social impact, health care is an immediate beneficiary,” said Soma Dhavala, director, machine learning at Wadhwani AI, an institute which studies how AI can be used to improve lives and livelihood.

Apollo has launched many AI initiatives. AI-CVD, for instance, is an intelligent platform to predict heart attacks―designed specifically keeping Indians in mind. It is developing similar AI-based algorithms for diabetes, cancer and other non-communicable diseases, too. Besides CIM, it uses ProHealth platform, an AI-enabled personalised predictive health-risk assessor that uses data and functions to personally guide individuals to manage their own health.

“With adoption and implementation of AI in the health care sector, medical experts are now being able to deliver a customised and more precise health care services to the patients,” said Reddy. The AI initiatives at the hospital chain range from better diagnosis and better monitoring (smart in-patient room automation with AI-powered triaging system that continuously monitors the patient’s heart rate and respiratory rate) to even in radiology.

Many health care startups are looking at including predictive analysis, where a patient’s condition is compared both with data of hundreds of thousands of other similar patient records (available through cloud servers) and the patient’s own status, or predict whether the patient is likely to have a recurrence. Such techs are increasingly in demand in many countries where immediate re-admission of an insured patient comes with government penalties.

Of course, the age of ‘goodbye doc, hello app’ may still have to wait, but progress in medical AI has reached a stage where tech companies, in India and abroad, are investing big in coming up with tailor-made solutions. Harman, a subsidiary of Samsung, recently came up with two offerings―one, a media suite that can track a patient to alert the nursing station or doctor the moment the patient behaves ‘abnormally’, and second, an Intelligent Healthcare Platform (IHP), targeted at making players right from insurance companies and pharma majors to the hospitals and their ancillaries’ systems interoperable.

Built at its Bengaluru centre and made for the world, these new AI tools could reimagine patient care. “In reality, patient data is only based on info from one company, whereas you can get much better decision matrices if you bring in data from the patient’s records in other systems too,” said Jai Ganesh, chief technology officer of Harman.

The proliferation of AI in the medical sector depends on how well its appetite for data is satiated, though that could bring with it its own set of issues. “To some extent it will be scary if you don’t implement a lot of the control mechanisms,” said Ganesh. “Responsible AI is important, the model should be explainable, bias free, trustworthy and ethical. The data scientists and machine learning engineers should be trained on some of these aspects, rather than (just) bytes and coding.”

While patient care is one area where AI could work its magic, it could save billions of dollars in the drug industry. Bringing a new drug to the market is an expensive process, with the drug discovery alone costing a third of the total cost. With AI designing drugs for various purposes, pharma can substantially cut costs and the time required for drug discovery, said Mayank Mathur, academic director of the Institute of Data Science at Indian School of Business in Hyderabad.

The World Economic Forum predicted that AI could sort India’s acute shortage of doctors (64 for one lakh people compared with a global average of 150). The government’s Integrated Disease Surveillance Programme last year onward started using automated disease monitoring right down to local levels to filter and collate events of interest, giving advance notice to a possible infectious outbreak. It is now leveraged by the verification cell of the National Centre for Disease Control.

Medical science could gain more once AI technologies evolve and advance. “AI techniques are being employed to solve complex problems such as predicting protein properties, designing molecules, optimising synthetic routes, and visualising protein-drug interactions and protein-protein interactions,” said Soharab Hossain Shaikh, assistant professor at BMU. “In the days to come, we may witness the integration of AI with tissue engineering, promoting bone regeneration and advancing the field of regenerative medicine.”