Interactive visualization of retinal astrocyte images

Analyzing high-resolution images of astrocytes is important in understanding the diseases, such as glaucoma and retinal detachment, to which astrocytes are known to become reactive. This is challenging because the cells are small, homogeneous, and closely packed. We propose an interactive visualization system designed for such images. Our system employs a probabilistic segmentation algorithm to help distinguish between cells. Design decisions on visualization and interactions were made based on the needs of the scientists, resulting in a visualization that shows the details of individual cells in the context of a large image mosaic. Our interactive system brings patterns of information to the surface, conveys uncertainty, and serves as a tool for astrocyte research.