Tomosynthesis increases mammography interpretation time

The addition of tomosynthesis to mammography causes an increase in interpretation time of images from screening examinations in comparison with the time it takes to interpret images from conventional digital mammography, according to a study published in the January 2014 issue of Radiology.

Digital breast tomosynthesis, which was approved by the FDA in 2011 for clinical use, has been shown to reduce diagnostic challenges, provide better visualization of breast tissues, improve lesion conspicuity, and more accurately characterize lesion margins. When combined with conventional digital mammography, tomosynthesis has been shown to improve performance, reduce recall rates, improve diagnostic accuracy, and increase cancer detection rates.

“Given these interpretive advantages, many centers across the United States and the world are now adopting combined tomosynthesis and mammography for use in both screening and diagnostic settings,” wrote the study’s lead author, Pragya A. Dang, MD, of Massachusetts General Hospital in Boston, and colleagues.

However, findings from some smaller studies have indicated that radiologist interpretation time may increase with the combined screening modalities, meaning workflow and resource management would be affected. Dang and colleagues thus decided to determine the effect of implementing a screening tomosynthesis program on real-world clinical performance by quantifying the differences in image interpretation times for conventional screening mammography and combined screening tomosynthesis and mammography screening.

During the prospective study, ten radiologists read images from 3,665 screening digital mammography or combined screening examinations for one hour uninterrupted sessions. Images from the exams were interpreted in at least five sessions per radiologist per modality and the number of cases reported during each session was recorded for each reader. Each radiologists’ level of experience was correlated to the average number of cases reported per hour. The authors used analysis of variance to assess the number of studies interpreted per hour and a linear regression model to evaluate the correlation between breast imaging experience and time taken to interpret images from the two modalities.

Results revealed that the mean number of studies interpreted in an hour was 23.8 for combined tomosynthesis and mammography. The mean number for digital mammography alone was 34.0. An average of 10.2 fewer studies were interpreted per hour during combined tomosynthesis and mammography than digital mammography sessions. The mean interpretation time of images was 2.8 minutes for combined tomosynthesis and mammography and 1.9 minutes for digital mammography alone. As years of breast imaging experience increased amongst the radiologists, the overall additional time required to read images from the combined examinations decreased.

“This increase in interpretation time may be within acceptable limits, given the other associated benefits of the technique, such as increased cancer detection, reduced false-positive rates, and streamlined diagnostic workflow,” wrote the study’s authors. “However, this estimation of the differential interpretation time should prove extremely useful in preparing for the effect on radiologists’ workload and in planning for staffing requirements and resource allocation as this technology is implemented,” they concluded.