Cell shape recognition and segmentation in fluorescence microscopy images.

Abstract

Many cellular processes involve the translocation of proteins from the cytosolic region to the cortex of the cell and vice versa. The dynamics of such processes is typically investigated by fluorescence imaging of GFP-labeled versions of these proteins. Quantitative analysis of the resulting fluorescence images requires image segmentation procedures that identify the cell within the image and subsequently divide the cell into a cytosolic and a cortical region to monitor the temporal evolution of the fluorescence signals within these parts of the cell separately. Here, we present an image segmentation protocol that we have developed for this type of data analysis. It consists of noise reduction, normalization, and thresholding steps to generate masks that define the cytosolic and the cortical regions of a cell. Based on these masks, the desired fluorescence signals can be extracted from the confocal microscopy images.

Adaptive microfluidic gradient generator for quantitative chemotaxis experiments

“…The Matlab function bwlabel assigns numbers to all detected areas. Our algorithm 24 then chooses the area of interest based on a weighting function that considers the cell area and its distance from the center of the ROI. Finally, depending on the modus of operation, the position of the ROI or the microscope stage is adjusted in order to center the cell of interest within the ROI.…”