Fig. 5: 3D Segmentation accuracy evaluation using M. xanthus biofilm images (cell density = 36.2%, and SBR = 1.58). | Nature Communications

Fig. 5: 3D Segmentation accuracy evaluation using M. xanthus biofilm images (cell density = 36.2%, and SBR = 1.58).

From: Non-invasive single-cell morphometry in living bacterial biofilms

Fig. 5

a Maximum intensity projection of a 3D M. xanthus fluorescence image. Cells were labeled with membrane-intercalating dye, FM4-64. Similar images were obtained at N = 120 different time points. b Maximum intensity projection of the manually obtained 3D segmentation result. c Maximum intensity projection of a CNN-based 3D segmentation result after LCuts post-processing. Cells that can be matched with the GT are displayed in the same colors as GT or otherwise colored in white. d Segmentation accuracy of compared algorithms parameterized in terms of cell counting accuracy (y-axis) and IoU matching threshold (x-axis). Source data are provided as a Source data file for Fig. 5d.

Back to article page