Table 8 Confusion matrix for the classification performance using the Unet++ model (trained only with canine images) on the feline lymphoma images. The labels on the left indicate the class labels by pathologists’ majority classification, whereas the labels on the top represent the estimated labels by our workflow. The workflows’ accuracy on this dataset was 81.57%.

From: Histological classification of canine and feline lymphoma using a modular approach based on deep learning and advanced image processing