J-Invariant Volume Shuffle for Self-Supervised Cryo-Electron Tomogram Denoising on Single Noisy Volume

2024-11-22

arXiv.org

AI Summary

Low signal-to-noise ratio (SNR) in cryo-electron tomography (cryo-ET) severely limits the visualization fidelity of 3D macromolecular structures. Existing self-supervised denoising methods, when applied to single noisy tomograms, suffer from information loss and incomplete noise modeling. To address this, we propose JiBU-Net—the first joint-invariant blind-spot U-Net specifically designed for self-supervised denoising of individual cryo-ET volumes. JiBU-Net innovatively integrates sparse center-masked convolutions, dilated channel-wise attention, and voxel-level unshuffle/shuffle demixing-remixing mechanisms, enabling multi-scale statistical noise learning and structural prior preservation without clean ground-truth labels. Evaluated on real cryo-ET datasets, JiBU-Net consistently outperforms state-of-the-art supervised and self-supervised methods, achieving superior noise suppression and higher subcellular structural fidelity. This work establishes a new paradigm for high-fidelity, label-free 3D structural analysis in quantitative structural biology.

Technology Category

• Machine Learning: Unsupervised & Self-Supervised Learning
• Computer Vision: Medical and Biological Imaging
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

Cryo-Electron Tomography (Cryo-ET) enables detailed 3D visualization of cellular structures in near-native states but suffers from low signal-to-noise ratio due to imaging constraints. Traditional denoising methods and supervised learning approaches often struggle with complex noise patterns and the lack of paired datasets. Self-supervised methods, which utilize noisy input itself as a target, have been studied; however, existing Cryo-ET self-supervised denoising methods face significant challenges due to losing information during training and the learned incomplete noise patterns. In this paper, we propose a novel self-supervised learning model that denoises Cryo-ET volumetric images using a single noisy volume. Our method features a U-shape J-invariant blind spot network with sparse centrally masked convolutions, dilated channel attention blocks, and volume unshuffle/shuffle technique. The volume-unshuffle/shuffle technique expands receptive fields and utilizes multi-scale representations, significantly improving noise reduction and structural preservation. Experimental results demonstrate that our approach achieves superior performance compared to existing methods, advancing Cryo-ET data processing for structural biology research

Problem

• Denoises Cryo-ET volumetric images.
• Addresses low signal-to-noise ratio.
• Improves structural preservation in denoising.

Innovation

• J-invariant blind spot network
• Sparse centrally masked convolutions
• Volume unshuffle/shuffle technique