ReLink: Reliable Deep Learning for Microscopic Motion Tracking.
- 1School of Computer Science and Technology, Harbin Institute of Technology (Shenzhen), Shenzhen, Guangdong, China.
- 2Institute of Biopharmaceutical and Health Engineering, Tsinghua Shenzhen International Graduate School (SIGS), Tsinghua University, Shenzhen, Guangdong, China.
Abstract
Single-particle tracking (SPT) serves as a fundamental technique for deciphering complex microscopic dynamics in biological and physical systems by resolving trajectories of individual particles. However, linking localizations into trajectories remains challenging in realistic experiments, where crowding, blinking, and localization errors lead to linking ambiguity. Both statistical and deep-learning approaches have been explored in these conditions, yet reliable link selection that generalizes across varying data distributions remains elusive. Here, we propose ReLink, a general deep-learning framework for reliable trajectory linking in SPT. By integrating physical priors into geometric representations and leveraging confidence-based adaptive filtering at inference, ReLink enables automated selection of trustworthy linkages, thereby preserving the fidelity of downstream trajectory-based analyses. Across challenging simulated scenarios spanning broad experimental conditions and motion regimes, ReLink consistently outperforms representative statistical and deep learning-based trajectory linking methods in both edge-level and trajectory-level evaluation metrics, leading to markedly improved downstream inference on public benchmarks. We further validate ReLink on diverse real-world experiments, from free diffusion systems to crowded intracellular environments, demonstrating its practical significance for quantitative interpretation of microscopic dynamics and its promise for next-generation trustworthy and automated SPT.
Related
- A false sense of circularity: hidden variability in RNase R enzymes confounds reliable circRNA validation
- Reliable and Efficient Agentic Protein Design with Toolset Harnessing
- Benchmarking single-cell format conversion tools reveals interoperability gaps and introduces CrossCell for reliable cross-ecosystem data exchange.
- Microscopic collisions govern macroscopic aerotactic spreading through emergent transport coefficients
- In-column Synthesis and Atomic-resolution Electron Microscopic Analysis of Metastable Reactive Nitrides
BibTeX
@article{jiang2026relink,
title = {{ReLink: Reliable Deep Learning for Microscopic Motion Tracking.}},
author = {Yuan Jiang and Xiaochen Feng and Hongda Wei and Hao Sha and Hongpeng Wang and Min Zhang and Sanyang Han and Yongbing Zhang},
journal = {Advanced science (Weinheim, Baden-Wurttemberg, Germany)},
year = {2026},
month = oct,
doi = {10.1002/advs.78541},
url = {https://europepmc.org/article/MED/42952159},
note = {CC BY},
}
RIS
TY - JOUR
TI - ReLink: Reliable Deep Learning for Microscopic Motion Tracking.
AU - Yuan Jiang
AU - Xiaochen Feng
AU - Hongda Wei
AU - Hao Sha
AU - Hongpeng Wang
AU - Min Zhang
AU - Sanyang Han
AU - Yongbing Zhang
PY - 2026
DA - 2026/10/28
JO - Advanced science (Weinheim, Baden-Wurttemberg, Germany)
DO - 10.1002/advs.78541
UR - https://europepmc.org/article/MED/42952159
AB - Single-particle tracking (SPT) serves as a fundamental technique for deciphering complex microscopic dynamics in biological and physical systems by resolving trajectories of individual particles. However, linking localizations into trajectories remains challenging in realistic experiments, where crowding, blinking, and localization errors lead to linking ambiguity. Both statistical and deep-learning approaches have been explored in these conditions, yet reliable link selection that generalizes across varying data distributions remains elusive. Here, we propose ReLink, a general deep-learning framework for reliable trajectory linking in SPT. By integrating physical priors into geometric representations and leveraging confidence-based adaptive filtering at inference, ReLink enables automated selection of trustworthy linkages, thereby preserving the fidelity of downstream trajectory-based analyses. Across challenging simulated scenarios spanning broad experimental conditions and motion regimes, ReLink consistently outperforms representative statistical and deep learning-based trajectory linking methods in both edge-level and trajectory-level evaluation metrics, leading to markedly improved downstream inference on public benchmarks. We further validate ReLink on diverse real-world experiments, from free diffusion systems to crowded intracellular environments, demonstrating its practical significance for quantitative interpretation of microscopic dynamics and its promise for next-generation trustworthy and automated SPT.
SN - 2198-3844
ER -