Bayesian nonparametrics in multi-particle superresolved tracking

Superresolved localization with experimental means such as PALM and STORM has previously been used to probe static molecular structures relevant to neurodegenerative diseases by resolving particles below the diffraction limit. However, superresolved localization does not provide any insight regarding the underlying dynamical processes.

In the meantime, localizing diffraction-limited particles when their point spread functions (PSFs) overlap substantially is a quite difficult problem. Especially, for the case of dynamic particles, when their PSFs overlap, the history these particle tracks themselves allows us to deduce particle locations even when the distance between the particles falls below the diffraction limit.

In the most ideal single-particle tracking framework, one would like to obtain three types of information simultaneously: 1) particle localizations; 2) linking of particle localizations across the images; and 3) particle number determination.

Existing single-particle tracking algorithms treat these steps separately and, as a result, they are unable to simultaneously track many particles.

We have overcome the limitations of existing algorithms by employing a fully self-consistent Bayesian-nonparametrics framework where all aforementioned steps are treated simultaneously. Thereby, we are able to achieve for the first time superresolved tracking. That means tracking as many as 10 particles within a volume the size of an E. Coli even when the distance between the particles fall below the diffraction limit.