The use of computer-assisted video image analysis in the enumeration of immuno-stained cells in tissue sections.

A novel method of computer-assisted video image analysis (VIA) was used to determine the number of immunostained cells in tissue sections. This method permitted an accurate and objective quantification of cells of a particular phenotype. This enumeration was achieved by measuring the area stained by a test monoclonal antibody (such as the T cell marker, CD3) and comparing it with the area stained by a leukocyte common (LCA) monoclonal antibody (CD45). The proportion of T cells within the total leukocyte population in a particular tissue was then calculated. The differentiation of positive (stained) and negative (unstained) cells was uniformly maintained by setting the computer to detect a threshold for staining intensity. This enabled consistency to be maintained within a tissue section as well as between sections stained with the same antibody. In the present study, we determined the phenotype of leukocytes in colonic carcinomas by VIA and compared this with results obtained by normal visual analysis. The VIA method showed distinct advantages over normal visual analysis especially in sections which contained moderate numbers of stained cells.