Elsharawy KA, Toss MS, Abuelmaaty SR, et al. - Digital images obtained from hematoxylin and eosin stained sections from a large invasive breast cancer (IBC) cohort were classified into training (n = 400) and validation (n = 1,200) sets, which were used to assess the optimal scoring method of nucleoli in IBC and their prognostic importance, and improve the grading of BC by including the nucleolar score. In five field views, the quantification of nucleolar prominence was found to be a cost-effective and reproducible morphological trait that could prognosticate IBC behavior and offer an alternative to pleomorphism to enhance BC grading performance.