MCUa: Context +Uncertainty awareness for Diagnosis of Breast Cancer

Histology image classification plays a critical role in the early detection of cancer. Pathological diagnosis relies heavily on digitised histology slides, and Convolutional Neural Networks (CNNs) have proven to be highly successful in this area. However, tissue classification remains a challenge due to the large size of digitised samples and the lack of contextual information. To address this issue, we present a novel CNN model called Multi-level Context and Uncertainty aware (MCUa) dynamic deep learning ensemble model. The MCUa model comprises several multi-level context-aware models that learn the spatial dependency between image patches in a layer-wise manner.