Section: New Results

Spot localization for TMA image analysis

Participants : Nam-Hoai Nguyen, Charles Kervrann.

A very first task of TMA (Tissue MicroArray) image analysis is to accurately localize spots (separate tissue core) representing arrays of 512 x 512 pixels each, in very large images of several thousands of pixels. For this purpose, we have investigated a three-stage methodological approach. First, since tissue cores are separately assembled in array (grid structure). We started to design a graphical model to eliminate image defects due to the presence of dusts or the imperfection of TMA blocks fabrication. In the second stage, a wavelet-like transform is currently used to recognize interested features (spots) given the size of spots a prori. Third, we started to investigate the superpixel-based image representation (SLIC) [27] , [55] to handle very large images and biological details inside each spot.

Partners: V. Paveau (Innopys company)