Quantitative mapping of super-resolution imaging artifacts to improve accuracy of nanoscale measurements of mitochondrial morphology

A significant barrier to the quantitative assessment of subcellular structures using single-molecule localization microscopy (SMLM) is the large variance in imaging quality that arises from several factors related to fixation, labeling, and imaging. Hence, there is a need to quantitatively assess image quality before quantitatively evaluating the imaged nanoscale structures. We used NanoJ-Squirrel’s algorithm to quantify image quality using the Resolution-Scaled Pearson Coefficient (RSP). We found that the accuracy of measurements from SMLM images significantly decreased when the RSP value was below 0.8, which was true for both simulated and experimental SMLM images in examining mitochondrial morphology. When applied to our previously developed quantitative toolbox for assessing nanoscale changes in mitochondrial structure, we found that filtering using an RSP value of 0.8 significantly reduced the variance in mitochondrial measurements between images, thereby proving that RSP can be used to quantitatively assess image quality before acquiring quantitative measurements from 3D-SMLM data.