Surface Recognition and Parametrization of Three-Dimensional Images from 
 Fluorescence Microscopy of Living Cells
 Peter Kasson

 
          Many problems in biology involve the motion of proteins on the 
 surface of cells.  One way these motions can be monitored is by labeling 
 proteins of interest with fluorophores and then tracking motions via 
 time-resolved three-dimensional florescence microscopy.  This process 
 generates a series of volume datasets at different time points.  Most 
 investigations conducted to date have at this point employed qualitative 
 visual analyses (e.g. Krummel et al.).  However, to carry out quantitative 
 analysis of protein motions, we desire a surface-parametrized image for 
 each time point rather than a field-of-view volume.  Effecting such a 
 transformation involves several stages:  1)  cell identification, 2) cell 
 tracking, 3) cell surface recognition, and 4) cell surface 
 parametrization.  These are non-trivial tasks because the cells are 
 irregular objects that move and deform over the course of the 
 experiments.  In addition to the cell-surface motion of fluorescently 
 labeled proteins, other confounding physical processes are fluorophore 
 quenching, synthesis of new protein, and protein degradation.
          I propose to concentrate on the last two tasks in the process of 
 analysis enumerated above:  taking a volumetric image of a cell and 
 generating first a surface and second a transformation from 
 (x,y,z,intensity) space in the volume to (u,v,intensity) space on the 
 surface.  In this work I plan to adapt existing methods for surface 
 reconstruction described by Zhao et al. and then devise a method for 
 parametrization of the surfaces thus generated.
 
 
 References:
 Krummel MF, Sjaastad MD, Wulfing C, Davis MM.  Differential clustering of 
 CD4 and CD3zeta during T cell recognition.  Science. 2000 Aug 
 25;289(5483):1349-52.
 
 Zhao, H.-K., Osher, S. and Fedkiw, R., "Fast Surface Reconstruction and 
 Hole Filling using the Level Set Method", The 8th IEEE International 
 Conference on Computer Vision (in press).