Table 2: Testing napari-MM3 on external datasets.

Quality of segmentation masks produced by running napari-MM3 on a subset of published datasets from other groups [13,15,22,41]. As exact boundaries are difficult to determine by eye, we considered a cell to be correctly segmented if the Intersection over Union of the predicted mask and ground truth mask was greater than 0.6 (Methods). To evaluate the quality of the segmentation, we report the Jaccard index [42,43].

Dataset	Correctly segmented cells	False positives	False negatives	Jaccard index
Ollion et al. (BACMMAN) [15]	228	4	1	0.98
Lugagne et al. (DeLTA) [12,22]	247	22	1	0.92
Sachs et al. (molyso) [13]	247	4	0	0.98
Jug et al. (MoMA) [41]	80	0	0	1