- Color decomposition algorithm
- Robotic TMA Analysis and Management Module
- Distributed Telemicroscopy Module
- Intelligent Archiving Module
- Segmentation algorithm based on color active contour models and robust estimation
- Texton signatures
- Deformable co-registration algorithms
Color Decomposition Algorithm (back to top)
One great challenge that we faced was how to quantitate the staining characteristics of specimens when two or more stains are utilized. To accomplish this required the development of reliable color decomposition module. The first generation algorithm made use of a polar transformation and peak detection algorithm to carry out the analysis. Proof of concept of the color decomposition approach using breast cancer tissue arrays stained with anti-Smad antibodies was described in a joint publication by our laboratory.
Robotic TMA Analysis and Management Module (back to top)
Distributed Telemicroscopy Module (back to top)
Intelligent Archiving Module (back to top)
Segmentation algorithm based on color active contour models and robust estimation (back to top)
3-Channel Weighted Embedded Color GVF
Texton signatures (back to top)
Textons are defined as conspicuous repetitive local features that human perceive as being discriminative among textures. The computational model for textons, introduced by Leung and Malik, are cluster centers in a feature space which is generated in responses to a fixed set of filter banks.
Deformable co-registration algorithms (back to top)
The method was shown to provide sub-voxel accuracy for co-registering 3D datasets. We have already performed experiments comparing the performance of the newly developed algorithm with those achieved using the Insight Segmentation and Registration Toolkit (ITK) and the algorithms developed at the French national institute for research in computer science and control (INRIA). Shown here are two video clips of the differences images acquired while the co-registration algorithms brings them into alignment after they have been translated, rotated and scaled to determine break points.
Since these are difference images a totally black frame indicates a perfect match between the two 3D data sets.