A new method for the comprehensive automatic morphologic image analysis of Langerhans cells

Standard techniques used to study Langerhans cell (LC) populations are tedious and time-consuming. We developed a rapid and comprehensive method to evaluate LC using automatic digital image analysis. Human epidermal sheets were stained with OKT6 monoclonal antibody and evaluated on an IBAS 2000 computerized automatic digital image analysis system. The fields were imaged on a Photomicroscope I using a high-resolution monochrome video camera. Each field was digitized, normalized, and filtered. The LC were discriminated by density; each field was interactively edited to separate overlapping cells and to bridge small gaps in the dendrites. The discriminated binary image of the LC was measured for number of cells per field, total area of individual cells per field, and percent of the field occupied by LC. The cell bodies were then separately discriminated and subtracted from the thinned cell image to allow a separate measurement of number and length of dendrites per cell. Finally, the binary cell image in a reference section was dilated to occupy approximately the total field area. This degree of dilation was then applied to all subjects to give the "area of influence." This methodology provides a rapid and comprehensive tool for the evaluation of LC.