Spårning av nanopartiklar med iSCAT och maskininlärning

Abstract

This study explores the tracking of particles of size 100 nm from iSCAT microscopy in two and three dimensions with machine learning and algorithm-based tracking methods. The study compares the accuracy of particle position predictions in two dimensions between the LodeSTAR method and the Radial Variance Transform algorithm (RVT), how these predictions influence the creation of partic le trajectories with MAGIK, and how the diffusivities of particles depend on the chosen detection method. Furthermore, the three-dimensional LodeSTAR method is modified to train on synthesized images of particles in varying depths in order to gain vertical prediction capabilites. The accuracy of the LodeSTAR method in three dimensions is evaluated by comparing the diffusivity of predicted par ticle trajectories from experimental data with their theoretical values, as well as through the analysis of trajectories and their covariances in the vertical dimension. The results for the two-dimensional LodeSTAR model indicate that RVT – based on covariance analysis – yields detections that indicate Brownian motion of the particles, but LodeSTAR performs better in terms of the number of correctly predicted particles and tracking trajectories over longer time periods. Furthermore, comparisons of diffusivity, covariance and particle tracings for three-dimensional detections suggest that the LodeS TAR method extracts information about particle depth, but that the detections lack accuracy. It is suggested that improving the synthesized data to better capture the particle shape in a greater depth range would yield more accurate results. In summary, the LodeSTAR method shows promising future potential for tracking particles in three dimensions from iSCAT microscopy footage