7) An emerging identity crisis

One unfortunate side effect of radiology's evolution is that it can create a rift between innovators and the people whose jobs those innovations impact. Generally speaking, companies exhibiting at RSNA 2017, particularly those engaging in machine learning and artificial intelligence, were most comfortable describing how much efficiency their software can bring to health care by automating certain processes and eliminating human error and other inconsistencies.


Ask most of them what all this will mean for the radiologists who will be having tasks taken from them and you will hear an answer like this: These tools will not replace radiologists, they will only serve to support their work and perhaps free them up to do other things.

Clearly, only the most fatalistic radiologist can realistically imagine being displaced by a robot within the next decade or so, but if a radiologist's main job is to interpret images then it's hard to see how deep learning algorithms won't eventually develop better "eyes" than any human.

The industry may not arrive there for a long time but it will likely be an awkward road heading in that direction.

8) From reading images to managing information

C'mon, will the robots really take over? Of course not! Let's try and stay grounded here folks. Talk that AI will make its way into every facet of imaging and will somehow “replace” radiologists is overblown. Instead, AI will serve radiologists. In the opening RSNA session, three radiologists with a bent toward futurism counseled a capacity crowd of over 4,000 to reinvent themselves through AI technology, (aka: play the hand they've been dealt).


“[Radiologists] have to embrace disruption or be disrupted,” cautioned RSNA president Dr. Richard L. Ehman. “We must guide the process rather than having it imposed upon us.”

Business Affairs Homepage