A novel approach to the quantification of hepatic stellate cells in intravital fluorescence microscopy of the liver using a computerized image analysis system.

Hepatic stellate cells (HSC) are nonparenchymal liver cells which reside in the space of Disse within the hepatic microcirculatory unit. HSC can be distinguished using intravital fluorescent microscopy (IVFM) due to the autofluorescence from their intracellular vitamin A. Herein we report on a novel approach for the quantification of video-recorded rat HSC images acquired by IVFM (excitation 360 nm/emission 420 nm) by the combined use of the "Cell counting macro" and the "Measurement macros" in the NIH image software. The approach involved two major steps using (i) the "Cell counting macro" for automatic detection, threshold-setting, and generation of a binary image of the vitamin A autofluorescence in the HSC images and (ii) the "Compute percent black and white" command in the "Measurement macros" to automatically determine the HSC density (%), which was then expressed as percentage of the total area of vitamin A autofluorescence-associated sites per observation area. Comparing the vitamin A autofluorescence areas in the original and the binary fashion HSC images revealed that the "Cell counting macro" was an optimal option for the analysis of the low-magnification (x10 objective) HSC images, whereas this macro was not suitable for the analysis of the higher magnification (40x objective) HSC images unless modifications were made. Our analysis revealed that HSC represent approximately 4-5% of the total area of the liver surface. In analyzing the higher magnification HSC microfluorographs, the use of the original "Cell counting macro" resulted in a significant underestimation of HSC density (60% reduction, P < 0.01) when compared with those analyzed using our modified macro. This study represents the first report of an automatic and reliable approach to the intravital fluorescent microscopic quantification of HSC using a computer-NIH image analysis system.