2026-06-09 北京大学（PKU）
◆LargePNetは、大視野情報を活用するために大カーネル畳み込み（RepLKConv）とピラミッド型アーキテクチャを採用し、画像全体の統計情報を学習へ反映することで復元精度を向上させた。8種類の蛍光顕微鏡画像タスクで検証した結果、既存のCNNやTransformerベース手法よりPSNRを0.5～2dB改善し、大規模画像の推論速度もCNN比で約4倍、Transformer比で約20倍向上した。さらに、本技術により200nm分解能で最大30時間の長時間ライブセル観察や、1時間に及ぶ3色STED超解像イメージングを実現した。
◆研究は、計算科学と顕微鏡技術を融合した次世代バイオイメージング基盤として、細胞動態解析や生命科学研究の発展に大きく貢献する成果である。
Figure 1 Architecture and design principles of LargePNet. (a) Limitations of patch-based training and discrepancies in structural statistics between small patches and large images. (b) Overview of the LargePNet architecture. (c) Effective receptive field analysis. (d) Comparison between LargePNet and conventional training strategies.
＜関連情報＞
- https://newsen.pku.edu.cn/news_events/news/research/15572.html
- https://www.nature.com/articles/s41467-026-71278-2
広範囲の統計情報を集約する復元ニューラルネットワークを用いて蛍光イメージングの限界を押し広げる Pushing the limits of fluorescence imaging with a restoration neural network aggregating large-view statistics
Yiwei Hou,Shu Gao,Wei Ren,Yunzhe Fu,Meiqi Li & Peng Xi
Nature Communications Published:07 April 2026
DOI:https://doi.org/10.1038/s41467-026-71278-2
Abstract
Deep learning has demonstrated remarkable success in augmenting fluorescence imaging under photon-limited conditions. However, existing restoration networks are typically devised for training with augmented patches far smaller than the full-view raw data, an overlooked aspect that compromises fidelity and noise-resistance due to the loss of global statistics. To address this limitation, we propose a large-patch network (LargePNet), which synergizes the large effective receptive field provided by shallow ultra-large-kernel convolutions and the nonlinear representation capabilities of deep networks through scale separation. It effectively and efficiently leverages large-view global information for restoration. Directly trained with large-view images, LargePNet shows contrasting advantages over state-of-the-art small-patch networks, with 0.5-2 dB higher peak signal-to-noise ratio across eight representative restoration tasks, involving implementations for single-image, video, and volumetric fluorescence data. For full-view processing, LargePNet generally holds around 4-fold and 20-fold higher computational efficiency compared to advanced convolution-based and Transformer-based networks, respectively. The assistance of LargePNet helps achieve 30-hour-long fluorescence imaging to monitor cytoskeleton dynamics, and hour-long tri-color super-resolution imaging to investigate organelle interaction, showcasing its advancement in live-cell imaging.
