Skip to main content
. 2021 Jan 11;11:395. doi: 10.1038/s41598-020-79880-0

Figure 3.

Figure 3

Malignant breast mass correctly classified by the nomogram. (a) A 51-year-old woman in the validation cohort had a screening US-detected 0.9-cm breast mass assigned as Breast Imaging Reporting and Data System (BI-RADS) 4A. The mass was confirmed as grade 1 invasive ductal carcinoma (pT1N0M0, estrogen receptor positive, progesterone receptor negative, human epidermal growth factor receptor type 2 negative) by US-guided 14-gauge core needle biopsy and subsequent surgery. (b) The deep learning based computer-aided diagnosis (DL-CAD) software assigned irregular shape score of 0.90 and parallel orientation score of 0.12. (c) This mass has 185 points on the nomogram, which was higher than the cut-off of 114, requiring biopsy. The nomogram incorporating quantitative morphologic scores from DL-CAD software correctly identified the true cancer.