PIV as a Method for Quantifying Root Cell Growth and Particle Displacement in Confocal Images

Particle image velocimetry (PIV) quantifies displacement of patches of pixels between successive images. We evaluated PIV as a tool for microscopists by measuring displacements of cells and of a surrounding granular medium in confocal laser scanning microscopy images of Arabidopsis thaliana roots labeled with cell-membrane targeted green fluorescent protein. Excellent accuracy (e.g., displacement standard deviation <0.006 pixels) was obtained for root images that had undergone rigid digital translations of up to 40 pixels. Analysis of zoomed images showed that magnifications of up to 5% maintained good linear relations between PIV-predicted and actual displacements (r(2) > 0.83). Root mean squared error for these distorted images was 0.4-1.1 pixels, increasing at higher magnification factors. Cell growth and rhizosphere deformation were tracked with good temporal (e.g., 1-min interval) and spatial resolution, with PIV patches located on recognizable cell features being tracked more successfully. Appropriate choice of GFP-label was important to decrease small-scale biological noise due to intracellular motion. PIV of roots grown in stiff 2% versus 0.7% agar showed patterns of cell expansion consistent with physically impeded roots of other species. Roots in glass ballotini underwent rapid changes in growth direction on a timescale of minutes, associated with localized arching of ballotini. By tracking cell vertices, we monitored automatically cell length, width, and area every minute for 0.5 h for cells in different stages of development. In conclusion, PIV measured displacements successfully in images of living root cells and the external granular medium, revealing much potential for use by microscopists. Microsc. Res. Tech. 73:27-36, 2010. (C) 2009 Wiley-Liss, Inc.