Dataset for Establishing a Reference Focal Plane Using Machine Learning and Beads for Brightfield Imaging

This dataset consists of sets of images corresponding to the data sets 1-8 described in Table 1 in the manuscript "Establishing a Reference Focal Plane Using Machine Learning and Beads for Brightfield Imaging".Data sets from A2K contain two .zip folders: one with the .tiff images and one with the corresponding .txt file with live and dead cell concentration enumeration. The A2K instrument software collects 4 images per acquisition, and each of those images is passed through the A2K instrument's software algorithm which segments the live (green outline), dead (red outline), and debris (yellow outline) objects. Segmentation parameters are set by the user. This creates a total of 8 stored images per acquisition. When in proper focus and brightness, the V100 beads are segmented in green, appearing as live cells. In cases where the beads do not display the bright spot center (when out of focus or too dim) the software may segment the beads in red, as dead cells.Data sets from the Nikon contain .zip folders of .nd2 image stacks that can be opened with Image J.These image sets were used to develop the AI model to identify reference focal plane as described in the associated manuscript.