The Invisible Gorilla: 83% of Radiologists Missed a Gorilla in a Lung Scan

When a gorilla was slipped into lung images, 83% of radiologists did not notice it. That is the result of an experiment by Drew, Võ and Wolfe, published in Psychological Science in 2013.

In a Brain Mondo short video, I introduced this experiment in about 40 seconds. In this post I write about the details that did not fit in the video, a later study that tested the same thing with real findings instead of a gorilla, and the idea of reducing misses with systems rather than effort.

Video: Focus facts: The invisible gorilla (Japanese)

The video is in Japanese.

What inattentional blindness is

When people are focused on something else, they can miss an unexpected event right in front of them, however obvious it is. Psychology calls this “inattentional blindness”.

The study that made this phenomenon famous was by Simons and Chabris, published in Perception in 1999 under the title “Gorillas in our midst”. It looked at how often people working on a task while watching moving scenes would notice an unexpected object.

One thing it found is that whether people notice depends on how similar the unexpected object is to the things they are attending to, and on how hard the task is. Interestingly, the unexpected object was not easier to notice when it was close to the place people were attending to. The authors suggested that people attend to objects and events, not to positions in space.

Most of these experiments, though, were done with ordinary participants on unfamiliar tasks. What about experts who have spent years sharpening their ability to find small abnormalities in a particular kind of image? That is what the 2013 study set out to test. Its title: “The invisible gorilla strikes again”.

What the experiment was

The participants were 24 radiologists. They were asked to do something from their everyday work: look for small spots in the lung, called nodules. They went through lung CT images and searched for small spots that might be cancer.

In the last case shown, the researchers had inserted a picture of a gorilla. It was 48 times the size of the average nodule they were looking for. For eyes trained to find small spots, it was large enough that missing it should have been the hard part.

The researchers also recorded where the doctors looked, using eye tracking. That way they could check afterwards where each person’s eyes had been.

Result: 83% did not notice

Eighty-three percent of the radiologists did not see the gorilla. Of the 24, only 4 noticed it.

The eye tracking showed something more surprising. Most of the doctors who missed the gorilla had looked directly at where it was. Their eyes had landed on the gorilla. Even so, their minds had not registered seeing it.

The paper concludes that even expert searchers, working in their own field of expertise, are vulnerable to inattentional blindness.

This result shows how unreliable the idea “if you look properly, you’ll notice” can be. Looking and noticing are two different things.

Experts still did better than non-experts

Here is a detail the video did not mention. According to the abstract of a 2021 paper, the 2013 study also gave the same task to ordinary people who were not doctors, and they missed the gorilla at a higher rate.

So being an expert is not useless against inattentional blindness. It seems to help to some degree. Experts’ eyes take in more than non-experts’ eyes do. Even so, more than 80% of them missed it. Both things are true.

What about real findings?

A gorilla is something that should never be in a lung image. You could argue that the doctors missed it because they had no reason to look for it in the first place. A study by Williams and colleagues, published in Psychonomic Bulletin & Review in 2021, put this to the test.

In this study, 50 radiologists read CT images to look for lung cancer. The last case showed a large 9.1 cm breast mass (breast cancer) and swollen lymph nodes. Both are real findings that would matter for diagnosis.

While their attention was on searching for lung nodules, 66% of the doctors did not detect the breast cancer, and 30% did not detect the swollen lymph nodes. But in a follow-up study, when 30 radiologists were asked to search for a broader range of abnormalities, the misses fell to 3% and 10%.

On top of that, neither years of experience, performance on the lung nodule search, nor how they searched the images predicted who would miss. The authors conclude that perceptual expertise does not protect against inattentional blindness, even for unexpected findings within the expert’s own field.

What the study does and does not tell us

The 2013 experiment had only 24 people, and the gorilla appeared only in the last case. You cannot read 83% as the rate at which misses happen in real hospitals. Images prepared for a study and everyday clinical work are different conditions.

The study also does not show that doctors are bad at their jobs. That is why the video said a miss is not carelessness but how the brain works. Because attention narrows onto something, you can find a tiny spot. The price is that things outside that narrow focus are hard to notice, even when they are in view.

In the 2021 study, asking doctors to search broadly greatly reduced misses, which fits this view. Change what you are looking for, and what you notice changes too.

Using it in daily life

Here are three ways to apply this research to daily life.

• For important checks, borrow someone else’s eyes. When you read a document or code you wrote yourself, what you are looking for is set from the start. Someone who does not know the background may find what was outside your attention.
• Check again with a different purpose. After reading for typos, go back and look only at the numbers, then only at the dates. As in the 2021 study, changing what you search for changes what you notice.
• Do not treat misses as a matter of effort. “I’ll pay more attention” does not change how the brain narrows its attention. Building a system, such as a checklist or a fixed checking procedure, is more reliable.

Brain Mondo is a series of daily short videos about the brain, each based on a research paper. BrainDojo is an app for training calculation, memory, attention and more in a few minutes a day, and it includes training for holding and switching attention. It can also be a chance to notice where your attention goes, and where it tends to slip away.

References

• Drew T, Võ ML, Wolfe JM. The invisible gorilla strikes again: sustained inattentional blindness in expert observers. Psychol Sci. 2013;24(9):1848-1853.
• Simons DJ, Chabris CF. Gorillas in our midst: sustained inattentional blindness for dynamic events. Perception. 1999;28(9):1059-1074.
• Williams L, Carrigan A, Auffermann W, et al. The invisible breast cancer: experience does not protect against inattentional blindness to clinically relevant findings in radiology. Psychon Bull Rev. 2021;28(2):503-511.