Visualization: Interactive 3D Histology Visualization for Pathology

Histology, the medical analysis at microscopic level of tissue specimen from the human body, is a cornerstone in healthcare and medical research. Digitization of the microscopy images opens up exciting possibilities for visualization and image analysis. A particularly exciting area for research purposes is 3D histology. The reason is that it is then possible to study the microscopic 3D structure, which has not been possible before, and to correlate that structure with in vivo imaging findings of the same tissue (before surgery/biopsy), such as Computed Tomography (CT) or Magnetic Resonance (MR) imaging. Some research labs today have the capacity to produce 2D slides that are possible to merge into a volumetric data set. But there are many visualization and image analysis obstacles to overcome in order to work effectively with this type of data.

In this project novel tools for visualization of 3D histology are developed. This includes effective volume rendering as well as interactive feature exploration methods. One challenge is the amount of data. A single microscopy image can be 100,000 x 100,000 pixels, and includes features over very different scales. The data is further fairly unbalanced, since the resolution in-between slides is much lower than within slides, and the slides are initially not perfectly matched due to the plastic nature of the tissue.