Section: Application Domains

Computational methods for bioimage informatics

In cell and molecular biology [62] , new challenges arise to acquire a complete and quantified view from the scale of a “single” cell to the scale of a multi-cellular structure, within the whole organism. In the near future, image analysis will be central to the successful use of optical microscopy in post-genomics biology. Nevertheless, one major difficulty lies in correlating and/or fusing multi-modalities, now routinely used in biology laboratories: optical imaging (spinning-disk confocal, TIRF, SIM, PALM, STED, FLIM-FRET, MP, SPIM/DSLM), ionic imaging (NanoSIMS), atomic force imaging (AFM) and electron imaging (Cryo-EM, Tomo EM).

Moreover, in the emerging era of high-throughput microscopy (biochemical screens, cell-based screening), systematic and accurate correlation and analysis of these data cannot be performed manually, since the image sequences are composed of several hundred of 3D stacks. Consequently, data to manipulate range from few to tens of TeraBytes. From the experimental perspective, molecular (drugs, RNA interference), mechanical (micro-patterning...), and optical (FRAP, photoactivation, optogenetic, ...) functional modulations allow one to quantify the importance of molecular linkage into macrocomplexes within a single cell. We are now able to limit shape variability between cells during an exposed period [59] , [54] . Consequently, efficient storage, fast retrieval and secure sharing of microscopy images are crucial challenges. Even with high-speed computers, the processing step will considerably slow down the whole analysis process.

We propose to address several important issues in this area and to adapt the proposed methodologies and algorithms to face a deluge of data. Our goal is also to participate to the technical specifications of an image database with a built-in query system to annotate, retrieve, process and integrate analysis from different imaging modalities. The combination of complementary skills (image processing and analysis software, image data management) will yield a full integration of the image and data life-cycle, from image acquisition and analysis, to statistical analysis and mathematical modeling in systems biology.