Abstract
Strategies for Analysis of Whole Slide Images
By Dr. Michael Grunkin, CEO Visiopharm
April 15, 2010
Manufacturers of Whole Slide Scanners are generally not providing anything beyond rudimentary image analysis capabilities, if any at all.
Therefore many users find that they are unable to realize the expected increase in productivity and capability with the new and promising Whole Slide Imaging technology, because they are left with several unmet needs to:
- Develop, adapt, and use standardized protocols for analysis requirements beyond simple color de-convolution and thresholding / phase analysis. Many applications require the ability to incorporate and combine spectral, spatial, morphological, contextual, and relational information in order to reach sufficient throughput and reliability of computed end-points.
- Provide automated identification and delineation of regions of interest, for elimination of unnecessary manual steps in the analysis pipeline. This can often (but not always!) be achieved through use of image analysis capabilities across multiple scales.
- Implement strategies for complex analysis of entire tissue sections with a computational effort that can both be paid for and executed in finite time.
- Become independent of proprietary image formats: Usually, device manufacturers only provide software that handles their own slide format. This severely inhibits collaborative efforts involving digital slides from different sources and scanners.
This webinar will demonstrate how these challenges can be effectively met with modern research tools for Whole Slide Images.
|
