Data imbalance is a major problem that affects several machine learning (ML)\nalgorithms. Such a problem is troublesome because most of the ML algorithms\nattempt to optimize a loss function that does not take into account the data\nimbalance. Accordingly, the ML algorithm simply generates a trivial model that\nis biased toward predicting the most frequent class in the training data. In\nthe case of histopathologic images (HIs), both low-level and high-level data\naugmentation (DA) techniques still present performance issues when applied in\nthe presence of inter-patient variability; whence the model tends to learn\ncolor representations, which is related to the staining process. In this paper,\nwe propose a novel approach capable of not only augmenting HI dataset but also\ndistributing the inter-patient variability by means of image blending using the\nGaussian-Laplacian pyramid. The proposed approach consists of finding the\nGaussian pyramids of two images of different patients and finding the Laplacian\npyramids thereof. Afterwards, the left-half side and the right-half side of\ndifferent HIs are joined in each level of the Laplacian pyramid, and from the\njoint pyramids, the original image is reconstructed. This composition combines\nthe stain variation of two patients, avoiding that color differences mislead\nthe learning process. Experimental results on the BreakHis dataset have shown\npromising gains vis-a-vis the majority of DA techniques presented in the\nliterature.\n