Extended Data Fig. 5: Evaluating SACD performance in 2D QD525 labelled microtubule versus SOFI.
From: Enhanced detection of fluorescence fluctuations for high-throughput super-resolution imaging
(a) A representative COS-7 cell labelled with QD525 under the SD-confocal and the SACD with 20 frames configurations was shown at the bottom and top sides. (b) Magnified views from the white box in (a) under different configurations. First row: SD-confocal, pre deconvolution followed by autocorrelation (AC) with 20 frames and full SACD reconstruction with 20 frames. Second row: pure AC with 20 frames, pure AC with 1000 frames, and AC with 1000 frames followed by RL deconvolution. Third row: pure cross-correlation (XC) with 20 frames, pure XC with 1000 frames, and XC with 1000 frames followed by RL deconvolution. (c) Fourier transforms of raw SD-confocal image, pure AC with 20 frames, AC with 1000 frames followed by RL deconvolution, and full SACD reconstruction with 20 frames. (d) Averaged intensity traces from pixels of the white circle in (b) for raw and raw after pre-deconvolution. (e) FRC analysis of the reconstructed images shows resolutions of 317 nm (raw SD-confocal), 206 nm (AC with 20 frames), 171 nm (AC with 1000 frames followed by RL deconvolution), 121 nm (SACD with 20 frames), and 118 nm (SACD with 1000 frames) respectively. (f) Intensity profiles and multiple Gaussian fitting of raw SD-confocal and 20-frame SACD for the structures of microtubule filaments indicated by the white arrows in (b). The numbers indicate the distance between peaks. Details in Supplementary Note 5. Experiments were repeated ten times independently with similar results. Scale bars: (a) 5 μm; (b) 500 nm; (c) 10 μm.
