Time-Lapse Microscopy of Cycling RPE1 Cells with Brightfield and H2B Imaging, Annotated with Cell Cycle States from integrated FUCCI Intensities

We acquired a high-resolution time-lapse fluorescence microscopy dataset of dividing human Fucci RPE1 cells, capturing images every 5 minutes. Cell nuclei were segmented using a custom StarDist model (Weigert and Schmidt) on an additional nuclear histone marker (H2B) channel and tracked across frames with TrackMate (Tinevez et al.) from 72-hour recordings. Full cell cycle tracks were identified through K-Means clustering, and ground-truth labels were derived by normalizing the average Fucci intensities per nucleus and applying a logarithmic transformation.

The training dataset includes 5,188 complete cell cycle tracks (M-M), with each track containing two imaging modalities—Brightfield and H2B fluorescence—represented as 64×64 images centered on the nucleus. We also provide two independent test datasets: one with unperturbed control cells (358 complete tracks) and another with drug-treated cells exhibiting cell cycle abnormalities (73 complete tracks).

We also included the model weights in the models.zip file. 

Imaging metadata:

Images from four channels—Brightfield, H2B (far red), Cdt1 (red), and Geminin (green)—were acquired every 5 minutes using a PerkinElmer Operetta Microscope with a 20x/0.80 objective (wide-field microscopy). Four or nine tiles per well were captured for each channel, with a 15% overlap for subsequent stitching. In the images, 1 pixel equates to 0.5979761uM. The laser intensities and time of exposure for each channel are shown in the table below.