Dr. Patel has a special interest in machine and deep learning, mathematical modeling of biological processes, informatics, and 3D printing. He is Chair of Department of Radiology, Medical Director for 3D Imaging and Printing Laboratory for Geisinger and leads the informatics efforts in Radiology at Geisinger. Dr. Patel has published 90 research articles, review articles, textbook chapters, and abstracts and has given several invited lectures at regional, national and international conferences. He is boarded in Diagnostic Radiology, Interventional Radiology, and Clinical Informatics and is a Fellow of the Society of Interventional Radiology.

In recent years, ML and in particular deep learning has revolutionized the field of computer vision. In ImageNet competition, deep learning models are now outperforming humans in object detection and classification. In addition, we have an unprecedented opportunity to combine data from different sources – EMR, imaging, genomics and apply machine learning for integration. We believe that using large clinical grade, heterogenous data set is extremely valuable in generalizing and translating to clinical tools. Advances in algorithms, vast volumes/variety of digital data and compute capabilities have finally reached a point where we can begin to ask questions that we have never asked before; combining all the -ologies,

-omics with imaging will lead to insights we have not had before.

The need is urgent -

The increasing data volume is leading to data overload and therefore data waste.

We are reaching the limits of human capacity due to data overload.

In the US alone, shortage of 90K physicians by 2025 and 120 K physicians by 2030 is expected.

What all this means is that we need AI - to make sure we can actually take care of patients by making the providers more efficient and to help with mundane tasks and easy tasks so that the providers can focus on more complex tasks. The key area we are focusing on are automatic detection/measurement, interpret/integrate findings, and predict outcomes and suggest therapies. The predictive capabilities and therapy suggestion options will enable population management.

It is clear that what physicians do will change – to what extent and how fast remains to be seen. AI will:

AI will help us be better physicians.

AI has and will improve patient care.

Physicians should be more involved in its evolution (revolution)

Complexity in medicine probably still has surprises in store for AI