In conventional Nanoparticle Tracking Analysis (NTA) small particles such as EVs are visualized by an ultra-microscopy set up and analyzed for size distribution from their Brownian motion (translational diffusion). Generally, sub-populations with different behavior and appearance can be distinguished by the naked eye but are difficult to quantify and to discriminate.  We present the transient NTA (t-NTA) method, which is derived from the statistical learning theory t-SNE [1] and projects a multi-dimensional solution into a two-dimensional scatter plot.