Comparison of readers' detection of right-sided and left-sided breast cancers and microcalcifications.

INTRODUCTION: Aim of this study was to determine if there is a statistically and clinically significant difference in diagnostic performance (cancer diagnosis) and perceptual performance (microcalcification detection) when detecting left-sided or right-sided breast cancers and microcalcifications. METHODS: Eight radiologist readers (8-20 years experience in radiology, five current BreastScreen readers) read a set of 100 digital mammograms (23/100 had proven malignancies and 52/100 had confirmed microcalcifications) for three reads (random case order in each read). The same mammograms were presented on two reads, serving as the baseline reads. The data from these reads were used to calculate intra-observer variability (presented in an earlier study). The experimental read consisted of left-right mirror images of the original mammograms. In each read, the radiologists were requested to 'clear' or 'call-back' cases and to indicate if any microcalcifications (benign and malignant) were present on the mammograms. Reading conditions were standardised. RESULTS: Comparison of intra-reader performance difference for left-sided versus right-sided breast cancers and microcalcifications with intra-observer variability for breast cancer diagnosis and microcalcification detection, respectively, revealed no clinically significant difference between left-sided and right-sided detections. Per-case analysis showed more left-sided breast cancers and microcalcifications correctly detected. This left-right difference in detection did not reach statistical significance, P-value of 0.28 for cancer diagnosis and 0.74 for microcalcification detection. CONCLUSION: There is no statistically or clinically significant difference between left-sided and right-sided breast cancer diagnosis and microcalcification detection in a group of experienced radiologists. Individual reading patterns do not affect detection rates of left-sided and right-sided cancers and microcalcifications.

Intrareader variability in mammographic diagnostic and perceptual performance amongst experienced radiologists in Australia.

INTRODUCTION: The purpose of this study was, for a group of experienced radiologists, to identify the magnitude of and statistical significance of intrareader variability in mammographic diagnostic performance or cancer diagnosis and mammographic perceptual performance or microcalcification detection. METHODS: Eight radiologist readers (8-30 years experience in radiology, five current BreastScreen readers) read a set of 100 digital mammograms on two separate reads with random case orders. Twenty-three of the 100 had proven malignancies, and 52 of the 100 had confirmed microcalcifications. The same mammograms were presented for both reads. The radiologists were requested to clear or call back cases and to indicate if any benign and malignant microcalcifications were present on the mammograms. Reading conditions were standardised. RESULTS: Intrareader variability in accuracy was demonstrated to be between 0% and 6% for the diagnostic task of breast cancer diagnosis and between 0% and 16% for the perceptual task of microcalcification detection. Intrareader agreement in the group of readers was high; between 75% and 93% (κ=0.36-0.72) for cancer diagnosis and between 77.5% and 93% (κ=0.17-0.77) for microcalcification detection. There was no correlation between reader's experience in radiology or being a BreastScreen reader and level of intrareader variability in cancer diagnosis and microcalcification detection. CONCLUSION: There exists intrareader variability in diagnostic and perceptual performance. Despite this variability, intrareader agreement remains high.

Comparison of 3MP medical-grade to 1MP office-grade LCD monitors in mammographic diagnostic and perceptual performance.

INTRODUCTION: Picture archiving and communication systems images designed to be viewed on high-resolution medical-grade monitors are routinely viewed on office-grade monitors on the wards or at home. This study aimed to determine whether a statistically significant difference in diagnostic (cancer detection) and perceptual (microcalcification detection) performance exists between 3MP grade and 1MP office-grade monitors. METHODS: 3MP Dome medical-grade liquid crystal display (LCD) monitors (Planar, Beaverton, OR, USA) were compared to 1MP Dell office-grade LCD monitors (Dell Inc, Round Rock, TX, USA). Eight radiologists (reader experience 8-30 years) read the same set of 100 mammograms (23/100 with proven cancers and 52/100 with microcalcifications) presented in random order on three occasions separated by two time intervals of 12 weeks. Reads 1 and 3 utilised 3MP monitors and formed the baseline read. Read 2 utilised 1MP monitors and constituted the experimental read. Reading conditions were standardised. Readers were aware of which monitors they were using. Multivariate logistic regression analysis (to account for reader variability and monitor impact) was performed to assess for statistical significance. RESULTS: At α = 5%, confidence intervals analysis comparing the measured parameters between 1MP to 3MP monitors demonstrated no statistically significant difference in diagnostic and perceptual performance for the reader group. In cancer detection (the diagnostic task), reader accuracy remained high irrespective of monitor type. Regression analysis comparing performance with 1MP against 3MP monitors found P values of 0.693 and 0.324 for diagnostic and perceptual performance, respectively. CONCLUSION: There were no statistically and clinically significant differences between 3MP and 1MP monitors in mammographic diagnostic and perceptual performance. Comparable performance may be due to compensatory behaviour by readers.