Computational counting for a quantitative analysis of cells in histologically prepared brain sections

Abstract

We propose a reliable method for automatic counting of cells in brain sections for different antibodies (NeuN, Parvalbumin, GABA and c-Fos) and Nissl-staining. The images of stained sections are binarized by thresholding. Then regions are clustered using a common clustering algorithm. After choosing only proper sized clusters the detected cell-bodies are counted. The parameters of the algorithm are adjusted manually and remain constant for different probes. The computational cell counting method provides correct counting results, shown by a comparison of computational results and counts gained by human experimenters.