Fig. 5From: A predicted-loss based active learning approach for robust cancer pathology image analysis in the workplaceExamples of noisy and clean patches (a) noisy patches that occur during the imaging process, involving substances or objects that are irrelevant to the classification, or of no interest; b noisy patches where meaningful tissue imaging is hindered by scanner errors or the presence of other substances; c clean patches that contain sufficient information for an accurate classification by an oracleBack to article page