Teleradiology and Artificial Intelligence

I recently spoke at a conference in New Delhi, the superbly organized and well attended, SER-APCIO joint conference.

My contributions included a lecture on the intriguing topic of More than just convenience: Teleradiology beyond a 24/7 global office model

I also spoke on a panel discussion on the topic Machine learning and Dr. Computer ‐ Is it the end of the road for radiologists? which was very ably moderated by my brilliant and charming colleague Dr Anjali Agrawal.

Preparing for these sessions made me realize that there are some striking similarities between Teleradiology and AI

a) Both are Disruptive technologies

b) Both threaten the status quo and generate fear among those radiologists who see it as a threat rather than an ally.

c) Both can actually help radiologists when used judiciously and scientifically, in the manner that they are meant to be used.

d) Both can transform healthcare delivery – by increase reporting accuracy and speed, ie both offer both productivity and quality benefits.

e) Both bring major cost benefits to healthcare

f) Both involve analytics and processing of large amounts of data

g) Of course both are IT enabled and follow Moore’s Law principles

h) Both can be seen either as a threat to be feared or as an opportunity to be embraced… ultimately its all in the perception.

In the concluding question of the panel discussion I was asked what radiologists can do to keep themselves relevant in today’s world and I responded with the following:

1. We need to cccept that we are humans, and get tired, and make mistakes. We need to stop resisting assistance, either by computers or by other radiologists.

2. We need to change our focus from being “lesion detectors” to analysis, problem solving, communication and education.

3. We need to focus on developing relationships with referring clinicians, streamlining our processes, and lowering imaging costs.

In this manner will we radiologists remain relevant, and no man or computer can or will ever replace us.

Scroll to Top