Coordinate-conditioned Deconvolution for Scalable Spatially Varying High-Throughput Imaging

TL;DR

This paper introduces SV-CoDe, a scalable deep learning framework that uses coordinate-conditioned convolutions to achieve high-resolution, uniform reconstruction in spatially varying high-throughput microscopy, overcoming previous limitations in memory and training costs.

Contribution

SV-CoDe is the first scalable, physics-aware deep learning method that locally adapts reconstruction kernels for spatially varying blur, enabling efficient high-resolution imaging over large FOVs.

Findings

Outperforms prior methods in image quality with less model size and training data.

Successfully generalizes from simulations to real biological samples.

Achieves uniform high-resolution reconstruction across a 6.5 mm FOV.

Abstract

Wide-field fluorescence microscopy with compact optics often suffers from spatially varying blur due to field-dependent aberrations, vignetting, and sensor truncation, while finite sensor sampling imposes an inherent trade-off between field of view (FOV) and resolution. Computational Miniaturized Mesoscope (CM2) alleviate the sampling limit by multiplexing multiple sub-views onto a single sensor, but introduce view crosstalk and a highly ill-conditioned inverse problem compounded by spatially variant point spread functions (PSFs). Prior learning-based spatially varying (SV) reconstruction methods typically rely on global SV operators with fixed input sizes, resulting in memory and training costs that scale poorly with image dimensions. We propose SV-CoDe (Spatially Varying Coordinate-conditioned Deconvolution), a scalable deep learning framework that achieves uniform, high-resolution reconstruction across a 6.5 mm FOV. Unlike conventional methods, SV-CoDe employs coordinate-conditioned convolutions to locally adapt reconstruction kernels; this enables patch-based training that decouples parameter count from FOV size. SV-CoDe achieves the best image quality in both simulated and experimental measurements while requiring 10x less model size and 10x less training data than prior baselines. Trained purely on physics-based simulations, the network robustly generalizes to bead phantoms, weakly scattering brain slices, and freely moving C. elegans. SV-CoDe offers a scalable, physics-aware solution for correcting SV blur in compact optical systems and is readily extendable to a broad range of biomedical imaging applications.