With IoT we see the primary impact when patients are able to avoid visiting the hospitals – devices allowing patients to be checked at their own location and diagnostic data to be shared with physician, thereby providing the requisite care without putting the additional burden of physical visit. This is especially relevant in remote areas which don’t have hospitals nearby and a visit to see the doctor may take one full day, putting serious strain on a day-labourer. While wearable solutions doing constant monitoring are useful for everybody, low-cost version of such solutions is especially useful in the low resource and remote settings.

Apart from the patient usage, hospitals also benefit from the IoT based devices. For example, constant monitoring of newborn with data available to doctor on mobile phone (thereby eliminating the need for nurse to regularly check) allows a doctor to monitor many more newborns than would be feasible with regular vitals-monitoring devices that are currently used. Different kind of devices being used for fitness and exercise tracking allow users to focus on their physical well-being, thereby preventing them from falling sick altogether and remaining healthy for longer duration.

Outside of the direct patients’ engagement, the IoT devices are able to improve health care operations by streamlining inventory management, tracking equipment usage, and optimising hospital workflows. For example, smart tags can locate essential medical equipment quickly, saving time during emergencies.

While IoT devices provide the additional monitoring capabilities through their connectivity, AI is beginning to play much bigger role especially in fields like radiology, oncology, precision medicine and personalised treatment for complex cases. Today the AI models are able to create high quality advisory by analysing the MRI scans and X-Rays, thereby reducing the workload on pathologists and physicians. AI creates an advisory report for the doctor, indicating any anomaly that it has found and directing their attention to the specific area. It is not feasible or even desirable to remove humans from the treatment cycle, but AI advisory can definitely make it easier for health care professionals to treat their patients.

Apart from assisting doctors, AI is beginning to play increasingly bigger role in identifying molecules for specific medicines. Normal drug discovery cycle for specific problem which would start with a larger number of molecules, can be shortened when AI model analyses those molecules and after performing simulations is able to reduce starting count to a fraction, leading to shortening of drug cycle as well as saving significant cost for pharmaceutical companies.

As we look towards the future, apart from the growth of various solutions using IoT and AI separately, we also see strong convergence between IoT and AI leading to solutions where both are playing critical part. When the devices are collecting the regular data from patients, instead of just leaving it to the consulting physician the AI can perform its own analysis and create advisory report for both patients and physicians. AI can also raise immediate alarm and initiate the pre-defined actions like triggering the ambulance service, alerting the family members etc. AI integrated with MRI and X-ray machine can generate the advisory report as tests are being conducted, thereby reduce the interval between diagnostics and treatment start. AI enabled microscopes can send the high resolution image of sample along with advisory report to remote pathologist who can look at both and share recommendation. Such a solution will be especially useful in remote areas which lack the qualified doctors. ASHA workers can take the diagnostic devices directly to the patients in villages and provide on-site diagnosis. The potential scenarios are unlimited and are expected to substantially improve the health care deliveries.

Development and adoption of these solutions also brings the challenges and ethical considerations. Issues like AI hallucination, access to clinical data for training AI models, data quality and anonymisation to prevent the patients’ privacy, informed consent to the data access and regulatory requirements are just some of the challenges that need to be addressed before AI in health care starts reaching its full potential. Ultimately, we want a future where health care is not only smarter but also more connected, personalised, safe, accessible, affordable and sustainable.

This article is authored by Sudhanshu Mittal, head & director, technical solutions, Nasscom CoE.