Getting poster data...
E. Wait, M. Winter, C. Bjornsson, S. Goderie, S. Temple, and A. Cohen (Drexel University, Philadelphia, PA; Neural Stem Cell Institute, Rensselaer, NY)Quantifying the development of clones, or family trees, of proliferating cells from time lapse image sequences is a fundamentally important task in biology. There is a pressing need for software to analyze these image sequences. Biologists must also be able to quickly visualize and validate the results. We have developed tools that combine powerful visualization with image processing techniques (registration, segmentation, tracking, lineaging) to enable interactive data exploration and validation. Our tools overlay analysis results on the image data using a unique blend of scripting (MATLAB), parallel computation (CUDA), and graphics hardware (NVIDIA GTX), along with active-shutter stereoscopic visualization (DirectX). Any errors in the analysis are quickly corrected interactively. User-supplied corrections are automatically applied to fix related mistakes. The tools have been used to distinguish stem cell behaviors in 2-D in-vitro image sequences. In 3-D timelapse images these tools quantified the relationship of stem cells to surrounding blood vessels and have enabled interactive examination of entire neural structures formed from a montage of dozens of overlapping image stacks.