Tracking-free mean square displacement analysis of molecular dynamics

Here we describe a protocol for image-derived mean squared displacement (iMSD) analysis, a fluctuation-based technique to quantify the average spatiotemporal dynamics of subcellular structures-such as proteins, organelles and nanoparticles-in living cells. Unlike single-particle tracking, which relies on discrete object localization and trajectory reconstruction, iMSD computes the spatiotemporal autocorrelation function of fluorescence intensity fluctuations directly from standard time-lapse image stacks. This approach overcomes single-particle tracking limitations in high-density or low signal-to-noise environments, providing a robust ensemble average without trajectory linking. The protocol details the optimization of image acquisition parameters, the computation of the spatiotemporal correlation function and the Gaussian fitting of the resulting correlation peak variance to derive the iMSD curve. Expected results include the simultaneous extraction of diffusion coefficients, anomalous exponents and confinement dimensions, allowing users to distinguish between Brownian, superdiffusive and confined motion across different temporal scales. The method's temporal dynamic range depends on the imaging modality: standard confocal scanning resolves millisecond-scale dynamics typical of organelles and membrane proteins, whereas camera-based or line-scanning configurations are required to capture fast cytosolic diffusion. Uniquely, the method also yields information on the average size of the moving objects. The complete procedure takes ~4-6 h of active hands-on time for image acquisition and computational analysis. It requires basic experience in fluorescence microscopy and data processing.