In recent years, hyperspectral imaging (HSI) has been shown as a promising imaging modality to assist pathologists inthe diagnosis of histological samples. In this work, we present the use of HSI for discriminating between normal andtumor breast cancer cells. Our customized HSI system includes a hyperspectral (HS) push-broom camera, which isattached to a standard microscope, and home-made software system for the control of image acquisition. Our HSmicroscopic system works in the visible and near-infrared (VNIR) spectral range (400 - 1000 nm). Using this system, 112HS images were captured from histologic samples of human patients using 20× magnification. Cell-level annotationswere made by an expert pathologist in digitized slides and were then registered with the HS images. A deep learningneural network was developed for the HS image classification, which consists of nine 2D convolutional layers. Differentexperiments were designed to split the data into training, validation and testing sets. In all experiments, the training andthe testing set correspond to independent patients. The results show an area under the curve (AUCs) of more than 0.89for all the experiments. The combination of HSI and deep learning techniques can provide a useful tool to aid pathologistsin the automatic detection of cancer cells on digitized pathologic images.
展开▼
