Coordinate-based neural representations for computational adaptive optics in widefield microscopy

TL;DR

This paper introduces CoCoA, a self-supervised neural network method that estimates wavefront aberrations and reconstructs 3D structures in widefield microscopy without external training, enabling in vivo mouse brain imaging.

Contribution

The paper presents CoCoA, a novel self-supervised neural approach that jointly estimates wavefront aberrations and 3D structures from single 3D images, eliminating the need for external training datasets.

Findings

CoCoA accurately estimates wavefront aberrations in widefield microscopy.

It enables in vivo imaging of mouse brains with adaptive optics.

The method is broadly applicable to various microscopy modalities.

Abstract

Widefield microscopy is widely used for non-invasive imaging of biological structures at subcellular resolution. When applied to complex specimen, its image quality is degraded by sample-induced optical aberration. Adaptive optics can correct wavefront distortion and restore diffraction-limited resolution but require wavefront sensing and corrective devices, increasing system complexity and cost. Here, we describe a self-supervised machine learning algorithm, CoCoA, that performs joint wavefront estimation and three-dimensional structural information extraction from a single input 3D image stack without the need for external training dataset. We implemented CoCoA for widefield imaging of mouse brain tissues and validated its performance with direct-wavefront-sensing-based adaptive optics. Importantly, we systematically explored and quantitatively characterized the limiting factors of CoCoA's performance. Using CoCoA, we demonstrated the first in vivo widefield mouse brain imaging using machine-learning-based adaptive optics. Incorporating coordinate-based neural representations and a forward physics model, the self-supervised scheme of CoCoA should be applicable to microscopy modalities in general.