Multi-dimensional Image Analysis Methods for Modern Optical Microscopy

Professor Badri Roysam

ECSE Department, Rensselaer Polytechnic Institute

Jan 26, 2006

Abstract:

Modern optical microscopes, aided by a family of support technologies, have matured into a truly multi-dimensional imaging tool for conducting diverse biological investigations at the sub- cellular, cellular, and tissue levels. The widespread availability of confocal and multi-photon microscopes and high-NA objectives, have made high-resolution (axial and lateral) three-dimensional (3-D) imaging of multiple structures and functional markers routine. From a systems biology perspective, modern microscopy is valuable for its ability to record processes in the spatial context of intact tissue, unlike biochemical assays, gene arrays and flow cytometry, in which spatial information is unavoidably disrupted. A variety of complementary advances in biochemistry (e.g., conjugated quantum dots, fluorescent proteins), image pre-processing software, and high- throughput specimen preparation serve to magnify the capabilities of optical microscopy.

There is a compelling need for technologies to translate this wealth of imaging data into quantitative insight. Specifically, two types of needs exist: (i) analysis of complex data sets involving interactions among multiple structures and functional markers, and multiple imaging dimensions; and (ii) analysis of large batches of such images. Quantitative measurements are needed in contexts ranging from hypothesis testing, drug discovery, assay development, high- throughput assays, quantitative tissue engineering, cytomic mapping.

In this talk, Dr. Roysam will describe the FARSIGHT project at Rensselaer that is producing software technologies that address the above needs in a systematic manner. This project is specifically designed for the multi-dimensional imaging capabilities of modern optical microscopy, and the complex image analysis needs of contemporary cell/tissue biology.