Table 1: Performance metrics for napari-MM3.

Processing times were measured on an iMac with a 3.6 GHz 10-Core Intel Core i9 processor with 64 GB of RAM and an AMD Radeon Pro 5500 XT 8 GB GPU. Tensorflow was configured to use the AMD GPU according to [40]. The GPU was used in U-Net training and segmentation steps. The dataset analyzed is from [10] and consists of 26 GB of raw image data (12 hours, 262 time frames, 2 imaging planes, 34 FOVs, and ~35 growth channels per FOV). Note that while the Otsu segmentation method is slightly faster than the U-Net, it also requires a background subtraction step, such that the total runtimes of the two methods are comparable.

	Channel detection	Background subtraction	Segmentation (Otsu)	Segmentation (U-Net)	Tracking	Total (Otsu)	Total (U-Net)
Frame processing time	N/A	2 ms	4 ms	5.3 ms	N/A	N/A	N/A
Channel stack processing time (262 time frames)	N/A	0.54 sec	1.14 sec	1.4 sec	0.7 sec	3.1 sec	2.1 sec
FOV processing time (35 channels)	14.1 sec	17.5 sec	36.5 sec	46 sec	46.7 sec	2 min	1.7 min
Exp. processing time (26 GB, 34 FOVs, ~20,000 cells)	3.2 min	9.9 min	20.6 min	26 min	26.4 min	60 min	55 min