Localization of Ultra-dense Emitters with Neural Networks
This work addresses a bottleneck in SMLM for biological imaging by enabling higher emitter densities, which is incremental but enhances experimental usability and temporal resolution.
The paper tackles the problem of overlapping emitter images in Single-Molecule Localization Microscopy (SMLM) by introducing LUENN, a deep convolutional neural network that extends usable emitter densities by a factor of 6 to over 31 emitters per micrometer-squared, improving temporal resolution and localization precision.
Single-Molecule Localization Microscopy (SMLM) has expanded our ability to visualize subcellular structures but is limited in its temporal resolution. Increasing emitter density will improve temporal resolution, but current analysis algorithms struggle as emitter images significantly overlap. Here we present a deep convolutional neural network called LUENN which utilizes a unique architecture that rejects the isolated emitter assumption; it can smoothly accommodate emitters that range from completely isolated to co-located. This architecture, alongside an accurate estimator of location uncertainty, extends the range of usable emitter densities by a factor of 6 to over 31 emitters per micrometer-squared with reduced penalty to localization precision and improved temporal resolution. Apart from providing uncertainty estimation, the algorithm improves usability in laboratories by reducing imaging times and easing requirements for successful experiments.