Project Summary Digital scanning of tissue slides, including both hematoxylin and eosin (H&E)-stained and immunohistochemistry (IHC)-stained slides, is becoming a routine clinical procedure. Technological advances in imaging, computing and molecular profiling have enabled in-depth tissue characterization at single-cell resolution while retaining the cell spatial information and its histological context. The confluence of these developments has created unprecedented opportunities for studying the relationships among tumor morphology, molecular events, and clinical outcomes. However, there is a lack of computational tools that can fully utilize the comprehensive information in tissue images at the single-cell level. The overarching goal of this proposal is to develop iSEE-Cell (image-based Spatial pattern ExplorEr for Cells), a suite of informatics tools to enable image data analysis, spatial modeling and data integration at single-cell resolution. In order to achieve this goal, we have built a strong research team with complementary expertise in image analysis, machine learning, spatial modelling, single cell genomics, cancer pathology and software development. Specifically, we will: 1. Develop algorithms to classify different types of cells based on nucleus morphology, that will be applicable to all types of tissue images. 2. Develop a powerful image restoration tool and quality enhancer for restoring blurred regions, enhancing low resolution/magnification into high resolution, and normalizing staining colors. 3. Develop and integrate tissue image analysis, spatial modeling and visualization tools into the iSEE-Cell platform. We will engage users, including informaticians, oncologists, pathologists, surgeons and cancer biologists, in the process of algorithm and tool development to collect feedback for the proposed informatics tools. All proposed methods were motivated by real-world biological and clinical applications. If implemented successfully, the proposed study will facilitate users in studying the tumor microenvironment and in improving cancer risk assessment, diagnosis, and outcome prediction.