A Framework for Particle Detection in Low-SNR Fluorescence Live-Cell Images and It’s Application for Rapid Single Particle Tracking in Massive Datasets

New microscopy imaging modalities still pose the same major challenge of traditional fluorescent microscopy techniques of trying to balance image quality against the viability of the cellular sample being investigated. Factors such as photobleaching and phototoxicity limit the intensity and the exposure time used to image the sample. This low light exposure combined with the increasing need to acquire data at a much higher frame rate results in images with a much lower Signal to Noise Ratio (SNR). In combination with new camera technologies the size of the datasets produced to study cellular dynamics using these imaging modalities has also increased greatly to the extent where standard data handling techniques cannot cope.
We present a robust particle detection framework that allows for rapid and automated tracking building on previous work. Also presented are improvements to the existing single particle tracking framework which enable us to handle massive datasets. We demonstrate the implementation of the detection framework and improved single particle tracking in a biological application. We examine the movement of individual SNAP-25 molecules at the cell membrane. The data acquisition imaging was completed using PALM with TIRFM at 100 frames per second using a scientific CMOS detector.
Furthermore we will briefly introduce Stimulated Emission Depletion microscopy and discuss future work relating to the application of existing denoising and deconvolution techniques to STED microscopy.