The Effects of Disease Category on Diagnostic Problem Solving in Mammography

Research on diagnostic problem solving is devoted to understanding the cognitive factors underlying the superior performance exhibited by medical professionals in their domain. In this study we analyzed think-aloud protocols and video recordings of staff and resident radiologists as they diagnosed mammograms during an interactive problem solving task. Analyses revealed statistically significant differences in radiologists' usage of problem solving operators (PSOs) and several performance measures between benign and malignant breast disease cases. Results suggest that additional factors outside of expertise should be considered in order to better understand diagnostic problem solving in a medical domain. The effects of disease type on performance and diagnostic problem solving should guide future development of medical tutoring systems.