Multicolor localization microscopy by deep learning

TL;DR

This paper introduces a deep learning approach for multicolor localization microscopy that classifies emitter colors from grayscale images without additional optical components, improving efficiency and design flexibility.

Contribution

It demonstrates neural networks can classify emitter colors using standard microscopy setups and designs phase elements for enhanced spectral differentiation.

Findings

High classification accuracy for static and mobile emitters

Neural networks outperform traditional spectral classification methods

Design of phase-modulating elements improves color differentiation

Abstract

Deep learning has become an extremely effective tool for image classification and image restoration problems. Here, we apply deep learning to microscopy, and demonstrate how neural networks can exploit the chromatic dependence of the point-spread function to classify the colors of single emitters imaged on a grayscale camera. While existing single-molecule methods for spectral classification require additional optical elements in the emission path, e.g. spectral filters, prisms, or phase masks, our neural net correctly identifies static as well as mobile emitters with high efficiency using a standard, unmodified single-channel configuration. Furthermore, we demonstrate how deep learning can be used to design phase-modulating elements that, when implemented into the imaging path, result in further improved color differentiation between species.