What do experts look at and what do experts find when reading mammograms?

Area of Science:

• Medical Imaging
• Radiology
• Cognitive Science

Background:

• Radiologists may miss clinically significant findings in mammograms.
• Eye tracking offers a method to investigate the causes of these reporting errors.

Purpose of the Study:

• To investigate the role of attention and eye movements in mammography errors.
• To understand why radiologists fail to report visible, significant findings.

Main Methods:

• Eye movements of 17 radiologists were tracked while viewing 80 mammograms.
• Errors were classified using the Kundel et al. taxonomy: search, recognition, and decision errors.
• Different definitions of the useful field of view (UFOV) were explored.

Main Results:

• Error proportions were 25% search, 25% recognition, and 50% decision errors.
• Few differences were observed between experts and residents in accuracy or eye movement metrics.
• The probability of eyes moving to the target was low, even when close, indicating attention deployment failure.

Conclusions:

• Mammography errors often stem from a failure in attention deployment, not solely from not fixating on significant locations.
• Observers do not fully process all information within their useful field of view.
• The study highlights the importance of attentional processes in diagnostic accuracy.