Fig. 3: The performance of Alpha-LFM. | Nature Communications

Fig. 3: The performance of Alpha-LFM.

From: Adaptive-learning physics-assisted light-field microscopy enables day-long and millisecond-scale super-resolution imaging of 3D subcellular dynamics

Fig. 3

a The volume rendering of the large-scale mitochondrial dynamics in dozens of live U2OS cells reconstructed by Alpha-Net within a large FOV of 220 × 220 × 10 μm3 using a 60×/1.3 NA objective. Scale bar, 20 μm. b The volume rendering of the reconstructed results of mitochondrial outer membranes in live U2OS cells by Alpha-Net, VCD-Net, and VS-LFD. Insets show the magnified volume renderings of the ROI indicated by the blue dotted box. Scale bar, 5 μm. c Decorrelation analysis quantifying spatial and axial resolution of VS-LFD, VCD-Net, and Alpha-Net. n = 20 volumes were analyzed at both xy and xz planes. The center line represents the median, the box limits represent the lower and upper quartiles, and the whiskers represent the min and max values. d Time-lapse 3D visualization captures the rapid morphological transformations of the mitochondrial outer membrane occurred at milliseconds timescale, illustrating both the processes of mitochondrial fission and fusion. Scale bar, 2 μm. e Comparisons of the photobleaching rates between Alpha-LFM, LSFM, Airyscan, and 3D SIM. f Max intensity projection of images of lysosomes in live U2OS cells imaged via 3D SIM (30 s every volume for a whole cell), Airyscan (4 min every volume for a whole cell), LSFM (4 s every volume for a whole cell), and Alpha-LFM (3 ms every volume for at least a whole cell). Scale bar, 2 μm. g The comparisons of the performance in resolution and volumetric imaging speed between Airyscan, iSIM, LLSM-SIM, FLFM, sLFM, VCD-LFM, and Alpha-LFM.

Back to article page