Fig. 1

Nuclei Segmentation: A Raw ‘.czi’ image of neuronal cultures. B Background filtered image of the raw ‘.czi’ image. We used the MATLAB function ‘imhmin’ (with 2 times the standard deviation of image as threshold) to suppress the signal derived from the neurites. C Segmented (using ‘Region Growing’) binary image of neuronal cell bodies. D Final watershed segmentation of ROIs from the binary image (inserts of ROI 13 and 14). Watershed was performed for better extraction of individual ROIs that are spatially in close proximity. We adjusted parameters (ROI size and eccentricity of ROI) to ensure that unintended splits and mergers were minimal. E ROIs (e.g., 4, 7, 11) with total pixel sizes below the size threshold are excluded from the final segmentation. F ROIs with an eccentricity > 0.99, indicating a line, and ROIs on the image border (i.e., 31) are excluded