This paper presents a deep neural network (DNN) approach for detecting mitosis in breast cancer histology images. The method uses a supervised DNN, specifically a max-pooling convolutional neural network (CNN), to classify each pixel in the image. The network is trained to distinguish patches containing a mitotic nucleus from those without. After training, the network is applied to a sliding window of the image, and the results are postprocessed to detect mitotic nuclei. The approach outperformed other methods in the ICPR 2012 mitosis detection competition. The challenge of mitosis detection lies in the similarity between mitotic and non-mitotic nuclei, which are both dark-blue spots. Mitotic nuclei are difficult to distinguish from non-mitotic ones, especially in later stages of mitosis when the nucleus may appear to split into two dark-blue spots. The DNN approach uses raw RGB data from square patches centered on each pixel to classify whether the pixel is part of a mitotic nucleus. The DNN is trained on a dataset of breast cancer histology images, with each pixel labeled as mitosis or non-mitosis based on its proximity to a known mitotic nucleus. The DNN is trained using a large dataset of images, with each pixel labeled as mitosis or non-mitosis. The network is then tested on a separate dataset, and its performance is evaluated using metrics such as recall, precision, and F1 score. The method is rotationally invariant, meaning it can detect mitotic nuclei regardless of their orientation in the image. To improve performance, the network is trained on multiple variations of the input image, including rotations and mirror images. The approach was tested on a publicly available dataset of breast cancer histology images. The results showed that the DNN approach outperformed other methods in terms of detection accuracy. The method is efficient, requiring only a few minutes to process a 4MPixel image on a standard laptop. The approach has the potential to be used in clinical practice for automated mitosis detection in breast cancer histology images.This paper presents a deep neural network (DNN) approach for detecting mitosis in breast cancer histology images. The method uses a supervised DNN, specifically a max-pooling convolutional neural network (CNN), to classify each pixel in the image. The network is trained to distinguish patches containing a mitotic nucleus from those without. After training, the network is applied to a sliding window of the image, and the results are postprocessed to detect mitotic nuclei. The approach outperformed other methods in the ICPR 2012 mitosis detection competition. The challenge of mitosis detection lies in the similarity between mitotic and non-mitotic nuclei, which are both dark-blue spots. Mitotic nuclei are difficult to distinguish from non-mitotic ones, especially in later stages of mitosis when the nucleus may appear to split into two dark-blue spots. The DNN approach uses raw RGB data from square patches centered on each pixel to classify whether the pixel is part of a mitotic nucleus. The DNN is trained on a dataset of breast cancer histology images, with each pixel labeled as mitosis or non-mitosis based on its proximity to a known mitotic nucleus. The DNN is trained using a large dataset of images, with each pixel labeled as mitosis or non-mitosis. The network is then tested on a separate dataset, and its performance is evaluated using metrics such as recall, precision, and F1 score. The method is rotationally invariant, meaning it can detect mitotic nuclei regardless of their orientation in the image. To improve performance, the network is trained on multiple variations of the input image, including rotations and mirror images. The approach was tested on a publicly available dataset of breast cancer histology images. The results showed that the DNN approach outperformed other methods in terms of detection accuracy. The method is efficient, requiring only a few minutes to process a 4MPixel image on a standard laptop. The approach has the potential to be used in clinical practice for automated mitosis detection in breast cancer histology images.