Fig. 4From: Deep learning networks find unique mammographic differences in previous negative mammograms between interval and screen-detected cancers: a case-case studySaliency maps of sample screen-detected and interval images (both correctly classified). For each row, the pseudo-presentation images are shown (left) along with the saliency map (middle) that highlights the pixels that had above a 50% weight in classifying the image in its respective category (i.e. first row saliency map highlights weights that push towards decision of classifying as screen-detected decision). At right, the images are overlaidBack to article page