Suman Chakraborty - Digital Health Symposium and Roundtable

Abstract: As healthcare providers continue to strive towards improving the patient outcomes irrespective of the barriers in affordability and accessibility, the use of Artificial Intelligence (AI) is becoming progressively more ominous in the healthcare system, ushering new hopes in several spheres ranging from underserved community centric primary healthcare to the super-specialized critical care at the resourced medical settings. However, the overuse or abuse of AI in healthcare is often criticised with a consideration that providing quality healthcare is not merely crunching of patient data and performing analytics on the top of it, so that the paradigm of relying too much on the machine’s ability to assess data and provide treatment plans seems unfeasible. Further, because of the sensitivity of healthcare from a holistic perspective including ethnic and cultural aspects and social acceptance, replacing human intelligence altogether appears an unrealistic ambition. In reality, rather, the dissemination of quality healthcare involves strong human-centric and personalized aspects including the use of “doctor’s wisdom” in a situation-specific manner that cannot be readily generalized via the established paradigm of AI. For health systems looking to leverage the advancements in e-health albeit within the boundary conditions of human-centric patient-care, the new horizon is the paradigm of intelligence augmented (IA), which includes AI but complements the same with different human intelligence elements in its approach as a decisive proposition. From a clinical decision making perspective, whereas AI is often portrayed to substitute the role of a doctor by undertaking the same assignment in an automated manner, IA acts as only a central tool to assist the healthcare professional, physician or policy decision maker on the hot-seat. Thus, instead of replacing human intelligence, IA emphasizes to build upon the same. In this talk, the aim to is to elucidate the framework of IA with three illustrative examples: (a) providing clinical decision support to doctors for patients under acute respiratory distress, and (b) cardiovascular decision support with a balance of physical device and AI tools, and (c) oral cancer and pre-cancer decision support using a cyber-physical technology, as developed by the speaker’s group. Further advancements to these approaches are envisaged by adapting several other technology elements, for example, the use of natural language processing, not only to transcribe patient-provider conversations during online consultation, but to assess and analyze the most salient points of the interaction for further attention as a post-processing measure. It is further emphasized that inclusion of IA with wearables and video-analytics can be a game changer for virtual care, specifically when it comes to remote patient monitoring. While each of the elements of IA as discussed herein by now experienced varied success, it is substantially more certain that when combined and integrated into a comprehensive platform, this can approach in a more pragmatic way to realize the central goal of ‘better care at lower costs’ that remained paradoxical thus far. A futuristic foundation of this is the development of a dynamically updated ‘patient-centric’ and ‘personalized’ digital health “twin” which, with the aid of IA, may trace and intervene the health trends and trajectories for the individuals as well as suggest personalized steps to better health. These tools can in practice help with so-called routine tasks as well as more delicate personalized aspects of diagnostics and treatment for quality healthcare under one umbrella. Without due preparedness for the IA transformation, the healthcare industry not- approaching to get aligned with the same may possibly be under the threat of being suddenly out-smarted and out-maneuvered: not by a human competitor, but by the combination of a ‘learned and wise human expert’ and a ‘dumb learning machine’.