The Washington PostDemocracy Dies in Darkness

Study: Most radiologists don’t notice a gorilla in a CT scan

It is, perhaps, a movie instantly familiar to anyone who has taken an introductory psychology class: The gorilla who walks through a group of kids passing a basketball between them and, for many, goes unnoticed.

The "invisible gorilla" experiment, as its come to be known, is meant to demonstrate inattentional blindness. When asked to focus on one particular factor—in this case, how many times students pass a ball between each other—other stimuli will go completely unnoticed. In this case, that's the man in a gorilla suit.

Turns out, this isn't just true for psychology students: A recent study found that a vast majority of radiologists failed to notice a very large gorilla that researchers added to a medical scan.

Trafton Drew, Melissa L. H. Vo and Jeremy M. Wolfe had 24 radiologists look at CT scans of lungs, scanning them lung nodules, which could be cancerous. Each CT scan had an average of 10 nodules and the radiologists were asked to click on them. The last of the five CT scans had something unexpected turn up on the lung: A gorilla. You can see him here, waving his arms at the top of the right lung.
gorillaMost radiologists, however, did not see him. When asked "Did you see a gorilla on the final trial?" 20 of the 24 radiologists tested said they did not. It wasn't for lack of looking. As the researchers write, in
a forthcoming paper in Psychological Science, "eye-tracking revealed that, of the 20 radiologists who did not report the gorilla, 12 looked directly at the gorilla’s location when it was visible."

On the more positive side, the radiologists did perform better than the untrained eye: When the researchers ran a group of non-doctors through the experiment, no one noticed the gorilla on the scan.

"Why do radiologists sometimes fail to detect such large anomalies? Of course, as is critical in all IB [inattentional blindness] demonstrations, the radiologists were not looking for this unexpected stimulus," they write. Even experts "operating in their domain of expertise, are vulnerable to inattentional blindness."