Abstract

To assess the usefulness of the lymph node imprint technique in quantitating cell types a comparative study using cytocentrifuge smears as controls was undertaken. The cell dimensions and distribution of cell types were compared in 12 patients with non-Hodgkin's lymphoma (NHL) using semi-automated image analysis. The results show that the imprint procedure introduces selective bias in favour of certain cell types; moreover, certain cell measurements, such as nuclear shape, are easier to see on cytocentrifuge smears. The imprint technique compares unfavourably with cytocentrifuge smears when precise quantitation of cell types is required.