A sliding-window data aggregation method for super-resolution imaging of live cells

Super resolution localization microscopy (SRLM) techniques such as STORM and PALM overcome the ~200nm diffraction limit of conventional light microscopy by randomly activating separate fluorophores over time and computationally aggregating their nanometer resolution detected locations for image reconstruction. However, a basic limitation of current SRLM approaches for live cell imaging is their low temporal resolution due to motion blur, which arises if image objects move during image acquisition of the substantial number of raw images required for constructing the super-resolution image for a given time point. To overcome this limitation, we propose a sliding-window data aggregation method, which exploits the temporal correlation between the collected fluorescence images to achieve significantly higher frame rate and therefore better temporal resolution than current approaches. Specifically, images within a sliding window are aligned so that locations of detected fluorophores within them are aggregated to accelerate image reconstruction for higher temporal resolution. We tested and validated our method using both simulated and real live cell STORM image data.