A study in the March issue of Radiology indicates that experience with digital breast tomosynthesis (DBT) may change the way breast imagers interpret 2D mammograms in subtle but significant ways. Researchers found higher recall rates in this scenario, but also a boost in cancer detection rates and positive predictive value (PPV).
The group from Brown University in Providence, RI, reported that although radiologists' recall rates for 2D full-field digital mammography (FFDM) went up after DBT experience -- from 6.8% to 7.9% -- the readers also found more cancers.
"The recall rate did go up, but it wasn't by much and was still within an acceptable range," corresponding author Dr. Ana Lourenco told AuntMinnie.com. "And it's reassuring that we found additional cancers. We were calling women back for things worthy of recall."
New tech's effect
Dr. Ana Lourenco from Brown University.
But the researchers wondered whether radiologists' experience with DBT would affect their screening recall and cancer detection rates for routine 2D digital mammography.
"We hypothesized that radiologist experience with DBT might alter ... [their] interpretive performance for digital mammography alone and perhaps decrease the digital mammography recall rate over time," the authors wrote.
DiPrete and colleagues reviewed data from 108,276 screening mammography exams performed at Rhode Island Hospital, also in Providence, between 2009 and 2014. Of these, 50,062 exams were performed before the hospital installed DBT and 58,214 were performed after it was installed. The group compared recall rate, cancer detection rate, PPV of positive screening result (PPV1), PPV of biopsy recommendation (PPV2), and PPV of biopsies performed (PPV3) for six radiologists during two time periods: 2009 to 2011 (before DBT) and 2012 to 2014 (after DBT).
The average recall rate among the six readers was 6.8% before DBT and 7.9% after it. Before experience with DBT, the recall rate increased only 0.01% for each year from 2009 to 2011 (not statistically significant). After experience with DBT, the recall rate increased 0.65% for each year from 2012 to 2014 (statistically significant). All other measures also increased after experience with DBT, the group found.
| Reader performance before and after DBT experience | |||
| Performance measure | Before DBT experience | After DBT experience | p-value |
| Recall rate | 6.8% | 7.9% | 0.031 |
| Cancer detection rate (per 1,000 exams) | 2.5 | 3.5 | 0.020 |
| PPV1 | 3.5% | 4.6% | 0.141 |
| PPV2 | 26.9% | 36.1% | 0.021 |
| PPV3 | 31.2% | 40% | 0.029 |
DBT's ability to visualize lesions more effectively than 2D mammography may be part of why radiologists' recall rates on 2D mammography increased, the researchers noted.
"We suspect that the increased recall rate for digital mammography alone after experience with DBT could be a result of decreased radiologist confidence when interpreting digital mammograms alone," the authors wrote. "The experience of seeing cancers at DBT when the concomitant digital mammographic findings are negative or subtle may have caused radiologists to change their threshold for recalling patients when interpreting digital mammograms only."
More sensitive
DBT experience heightens radiologists' sensitivity when reading 2D mammograms, Lourenco told AuntMinnie.com.
"We learn to see things better with tomosynthesis experience," she said. "The more you see on DBT, the more you wonder what's hiding on 2D. A 2D mammogram may look fine, but then you see the DBT image and there's the cancer. It's humbling."
