Microscopic differential interference contrast image processing by line integration (LID) and deconvolution

Abstract

Differential interference contrast (DIC) microscopy is the preferred imaging mode for studying live unstained cells and embryos. This is mainly attributed to the optical sectioning capability which makes DIC suitable for three-dimensional microscopy. However, DIC images are not convenient for standard computerized image interpretation. We present here a processing algorithm that converts DIC images into a form which is much more amenable for such analysis. The algorithm is based on computational inversion of the directional gradient performed optically by DIC, namely path integration (line integrated DIC, or LID). LID images relate to the refractive index of the specimen, and are inherently positive in value, thus many powerful algorithms developed for fluorescent images can be applied to them. The method is demonstrated here for identifying and tracking nuclei in developing zebrafish embryos, and for reconstructing the shape of live leukocytes.