From Whole Slide Tissues to Knowledge: Mapping Sub-cellular Morphology of Cancer

摘要: Digital pathology has made great strides in the past decade to create ability computationally extract rich information about cancer morphology with traditional image analysis and deep learning. High-resolution whole slide images of tissue samples can be analyzed quantitatively characterize cellular sub-cellular phenotypic imaging features. These features combined genomics clinical data used advance our understanding provide opportunities discovery, design, evaluation new treatment strategies. Researchers need reliable efficient algorithms software tools that support indexing, query, exploration vast quantities order maximize full potential digital research. In this paper we present a brief overview recent work done by group, as well others, systems.