Deep-STORM: super-resolution single-molecule microscopy by deep learning

TL;DR

Deep-STORM is a fast, accurate deep learning-based method for super-resolution microscopy that works without prior structural information, outperforming existing techniques in resolution and speed.

Contribution

It introduces a deep convolutional neural network approach for super-resolution microscopy that is trained on simulated or experimental data, achieving state-of-the-art results without prior shape information.

Findings

Achieves high-resolution images under noisy, high-density conditions

Significantly faster than existing super-resolution methods

Applicable to any blinking data-set without prior structural assumptions

Abstract

We present an ultra-fast, precise, parameter-free method, which we term Deep-STORM, for obtaining super-resolution images from stochastically-blinking emitters, such as fluorescent molecules used for localization microscopy. Deep-STORM uses a deep convolutional neural network that can be trained on simulated data or experimental measurements, both of which are demonstrated. The method achieves state-of-the-art resolution under challenging signal-to-noise conditions and high emitter densities, and is significantly faster than existing approaches. Additionally, no prior information on the shape of the underlying structure is required, making the method applicable to any blinking data-set. We validate our approach by super-resolution image reconstruction of simulated and experimentally obtained data.