The purpose was to study the ability of radiologists to detect breast cancers using 1-view breast tomosynthesis (BT) compared to 2-view digital mammography (DM) correlated with their experience in mammography. The patient population was enriched with difficult cases (89 abnormal and 96 normal/benign breasts). Eight breast radiologists with various experience levels in mammography interpreted the BT and DM image sets individually in a FROC study. Their performance was measured by the JAFROC figure-of-merit ( (non-parametric area under the AFROC curve) and analyzed as a function of experience level. The improvement was significant for the highly experienced radiologists, mean (θ BT-DM = 0.092; 95% CI: [0.023, 0.161] and for the experienced radiologists, mean (θ BT-DM = 0.094; 95% CI: [0.034, 0.149] while it was not for the less experienced radiologists, mean (θ BT-DM = 0.021: 95% CI: [-0.161, 0.202]. The results indicate that experience is necessary to achieve optimal performance in BT. © 2012 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Svahn, T., Lång, K., Andersson, I., & Zackrisson, S. (2012). Differences in radiologists’ experiences and performance in breast tomosynthesis. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7361 LNCS, pp. 377–385). https://doi.org/10.1007/978-3-642-31271-7_49
Mendeley helps you to discover research relevant for your work.