Segmentation of parasites for high-content screening using phase congruency and grayscale morphology

Abstract

Schistosomiasis is a parasitic disease with a global health impact second only to malaria. The World Health Organization has determined new therapies for schistosomiasis are urgently needed, however the causative parasite is refractory to high-throughput drug screening due to the need for a human expert to analyze the effects of putative drugs. Currently, there is no vision system capable of relieving this bottleneck with sufficient accuracy for the automated analysis of parasite phenotypes. We presented a region-based method with performance limited primarily by poor edge detection caused by body irregularities, groups of touching parasites and unpredictable effects of drug exposure. Towards ameliorating this difficulty, we propose an edge detector utilizing phase congruency and grayscale thinning. The detector can be used to impose the correct topology on a segmented image - an essential step towards accurate segmentation of parasites. © 2012 Springer-Verlag.