“…To circumvent this subjectivity, recent advances in imaging combined with machine learning have provided new avenues [103,104]. These include parametric machine-learning techniques that use quantitative nuclear morphometric information, such as nuclear size, shape, nucleus:cytoplasm ratio, and chromatin texture, for classifying histopathology images [105,106]. More recently, various nonparametric methods, including deep learning, have been applied for diagnosing various cancers, including breast, skin, and thyroid cancer, with high accuracies [107,108].…”