3D Digital Pathology Visualization
Histology, a subfield of pathology, deals with the medical analysis at microscopic level of tissue specimen from the human body. It is a cornerstone in health care and medical research. Digitization of the microscopy images opens new exciting possibilities. This includes moving from 2D image analysis to studying 3D structures and correlating them with in vivo imaging of the same tissue from CT or MR-imaging.
In this project tissue samples are studied on a scale of the cellular level. The aim of the project to develop efficient multi-scale visualizations of 3D stacks of collections of large-scale histology image. The colored images obtained by scanning of such tissue samples can exceed resolutions of 100,000 by 100,000 pixels at an optical magnification of 40x (>100 GiB per sample). The biggest challenges targeted include: The size of the datasets using volume rendering techniques; The design of effective transfer functions for RGB data in digital pathology; The development and evaluation of intuitive interfaces for the use by domain experts.
Digital Pathology Project website
The project builds upon a tight collaboration between computer scientists, pathologists, and clinical personnel.
Martin Falk, martin.falk at liu.se, website
- Martin Falk, Scientific Visualization Group, Linköping University
- Claes Lundström, CMIV, Linköping University and Sectra AB
- Darren Treanor, Leeds Teaching Hospitals NHS Trust, UK, and University of Leeds, UK, and Linköping University
- VINNOVA – DigiPat project
- Swedish e-Science Research Centre
- Excellence Center at Linköping and Lund in Information Technology (ELLIIT)
- Leeds Teaching Hospitals NHS Trust, UK, and University of Leeds, UK
- Karolinksa Institutet, Stockholm
Martin Falk, Darren Treanor, Derek Magee, Claes Lundström: 3D Histology—Interactive High-resolution Volume Rendering for Pathology. Nordic Symposium on Digital Pathology, Linköping, Sweden, 2015.