Abstract
Abstract: :
Purpose: The ConfoScan 3 clinical confocal microscope provides sharper images of cells in the stroma than the Tandem Scanning does. In this study we compared volumetric densities of cells determined from images recorded by these two instruments. Methods: Fifty corneas of 25 normal subjects were examined by confocal microscopy, first by using a Tandem Scanning confocal microscope (Tandem Scanning, Inc., Reston, VA), and then by using a ConfoScan 3 (Nidek Technologies America, Inc., Greensboro, NC). Two frames were selected from the mid–stroma (between 33% and 66% of the stromal depth) and bright objects, assumed to represent keratocytes, were counted in a known area of the frame. We limited our sample to the mid–stroma because motion artifacts with the ConfoScan 3 precluded us from precisely identifying the frames adjacent to the epithelium and endothelium. The effective depth of the sample volume represented by each frame was estimated from the number of consecutive frames and the corresponding distance that selected cells were visible and countable during a scan. The effective depth was the distance between the first and last frame that contained the same visible cell, plus the mean step distance per frame. Distances were determined from the frame rate and scan speed of the Tandem Scanning microscope, and from a z–axis encoder (prototype available for a limited number of exams) of the ConfoScan 3. Density was the number of visible cells in the sample area divided by the effective sample volume. Results: The effective focal depth of the Tandem Scanning microscope was 11.9 ± 2.6 µm (mean ± SD, n=32 cells from 5 scans), and of the ConfoScan 3 was 25.9 ± 7.1 µm (n=12 cells from 3 scans). Mean cell density at mid–stroma was 23,013 ± 4,420 cells/mm3 with the Tandem Scanning microscope and 23,996 ± 2,898 cells/mm3 with the ConfoScan 3. This difference was not significant (p=0.15, paired t–test; 90% chance of detecting a difference of 2,221 cells/mm3, if such a difference exists). Conclusions: In normal subjects, the ConfoScan 3 and the Tandem Scanning confocal microscopes indicate stromal cell densities that are not significantly different from each other. Estimates of cell density from both instruments require an accurate estimate of the effective depth of the sample volume, and this depth is approximately 2.2 times greater with the ConfoScan 3. This difference must be considered when comparing results from studies that use one instrument with results from studies that use the other.
Keywords: cornea: stroma and keratocytes • microscopy: confocal/tunneling • imaging/image analysis: clinical