3-D Image-to-Image Fusion in Lightsheet Microscopy by Two-Step Adversarial Network: Contribution to the FuseMyCells Challenge

TL;DR

This work focuses on fusing high-quality 3D volumes from a single lightsheet view to alleviate depth-related degradation and phototoxicity. It introduces a two-step 3D image-to-image fusion pipeline: a downsampled first pass to capture global context, followed by a patch-based high-resolution refinement guided by adversarial loss and a difference-weighting mechanism, all built on a 3D RUNet backbone. Results indicate that preserving global context and incorporating adversarial feedback improve fusion fidelity, with difference weighting reducing hallucinations, though computational cost is high and generalization to unseen data remains challenging. Overall, the approach advances 3D fusion in lightsheet microscopy and has potential to extend live imaging and multi-channel capabilities under photon-budget constraints.

Abstract

Lightsheet microscopy is a powerful 3-D imaging technique that addresses limitations of traditional optical and confocal microscopy but suffers from a low penetration depth and reduced image quality at greater depths. Multiview lightsheet microscopy improves 3-D resolution by combining multiple views but simultaneously increasing the complexity and the photon budget, leading to potential photobleaching and phototoxicity. The FuseMyCells challenge, organized in conjunction with the IEEE ISBI 2025 conference, aims to benchmark deep learning-based solutions for fusing high-quality 3-D volumes from single 3-D views, potentially simplifying procedures and conserving the photon budget. In this work, we propose a contribution to the FuseMyCells challenge based on a two-step procedure. The first step processes a downsampled version of the image to capture the entire region of interest, while the second step uses a patch-based approach for high-resolution inference, incorporating adversarial loss to enhance visual outcomes. This method addresses challenges related to high data resolution, the necessity of global context, and the preservation of high-frequency details. Experimental results demonstrate the effectiveness of our approach, highlighting its potential to improve 3-D image fusion quality and extend the capabilities of lightsheet microscopy. The average SSIM for the nucleus and membranes is greater than 0.85 and 0.91, respectively.

Figures (2)

• Figure 1: The pipeline of the proposed method.
• Figure 2: Exemplary outcomes of the proposed method.