FocAn: automated 3D analysis of DNA repair foci in image stacks acquired by confocal fluorescence microscopy

摘要

Flow chart demonstrating the main steps of the FocAn algorithm consisting of two independent components for nuclei ( ) and foci ( ) identification. In the first step (A), the raw image is normalized. Nucleus identification is then performed using a mean auto local threshold (ALT, B) followed by Gaussian blurring ( ). Together, these steps result in gradual signal separation of nuclear and cytosolic areas ( ), which is also illustrated in ( ). The green, red and blue lines in ( ) represent the intensity profiles of the corresponding colors in ( ), ( ) and ( ), respectively. After that, mid-gray ALT creates a binary image, shown in ( ). This is followed by watershed transformation for separation of overlapping nuclei, encircled with red lines in ( ). The foci identification process starts with Gaussian blurring of the normalized images followed by median ALT ( ). A 3D watershed transformation can be performed optionally, before finally the foci numbers per nucleus are determined ( )