Full prediction model could prevent 45.5 percent of false-positive recalls, 21.3 percent of benign biopsies for those with extremely dense breasts<\/p>\n<\/h3>\n
<\/b><\/p>\n
<\/b><\/p>\n
TUESDAY, Aug. 17, 2021 (HealthDay News) — For women with extremely dense breasts, prediction models based on magnetic resonance imaging (MRI) findings can reduce the false-positive first-round screening MRI rate, according to a study published online Aug. 17 in Radiology<\/em>.<\/p>\n Bianca M. den Dekker, M.D., from Utrecht University in the Netherlands, and colleagues used data from a randomized clinical trial that prospectively collected clinical characteristics and MRI findings in women with extremely dense breasts who had positive first-round MRI screening results after a normal screening mammography. In this secondary analysis, prediction models were built to distinguish true-positive from false-positive MRI screening findings.<\/p>\n The researchers found that 79 of the 454 women with a positive MRI result in a first supplemental MRI screening round were diagnosed with breast cancer (true-positives) and 375 had false-positive results. The full prediction model based on all collected clinical characteristics and MRI findings could have prevented 45.5 percent of false-positive recalls and 21.3 percent of benign biopsies, without missing any cancers (area under the receiver operating characteristic curve [AUC], 0.88). Comparable performance (AUC, 0.84) was seen for a model solely based on readily available MRI findings and age; this model could have prevented 35.5 and 13.0 percent of false-positive recalls and benign biopsies, respectively.<\/p>\n “Our prediction models may identify a substantial number of false-positives after first-round supplemental MRI screenings, reducing false-positive recalls and benign biopsies without missing any cancers,” den Dekker said in a statement. “This brings supplemental screening MRI for women with dense breasts one step closer to implementation.”<\/p>\n Several authors disclosed financial ties to biopharmaceutical companies, including Bayer Pharmaceuticals, which partially funded the study.<\/p>\n Abstract\/Full Text<\/a><\/p>\n Editorial (subscription or payment may be required)<\/a><\/p>\n <\/i><\/p>\n