Defocus and Low CNR Detection for Cell Tracking Applications

A growing number of screening applications require the automated monitoring of cell populations including cell segmentation, tracking, and measurement. Due to the highthroughput and low-dose nature of such screens, resulting images can suffer from microscope defocusing effects and low contrastto-noise ratios, which can lead to failed segmentation and tracking. We have previously presented general methods for cell segmentation and tracking that exploit the spatio-temporal nature of the task to constrain segmentation. Here we extend that work by developing algorithms using wavelet energy and distribution entropy for automatically detecting and removing defocused images and images with low contrast-to-noise (CNR) ratios before the segmentation and tracking steps. The results demonstrate the ability of the approach to effectively distinguish between normal images, defocused images, and those with lowCNR with an accuracy of greater than 95%.