@article{pmid41970592,

title = {Lightweight CycleGAN models for cross-modality image transformation and experimental quality assessment in fluorescence microscopy},

author = {Mohammad Soltaninezhad and Yashar Rouzbahani and Jhonatan Contreras and Francisco Paez Larios and Paul M Jordan and Oliver Werz and Rohan Chippalkatti and Daniel Kwaku Abankwa and Christian Eggeling and Thomas Bocklitz},

doi = {10.1364/BOE.578297},

issn = {2156-7085},

year  = {2026},

date = {2026-03-01},

urldate = {2026-03-01},

journal = {Biomed Opt Express},

volume = {17},

number = {3},

pages = {1476--1498},

abstract = {With the growing integration of artificial intelligence in scientific and medical applications, lightweight deep learning models have become increasingly important. These models offer substantial reductions in memory usage and computational time. Given that GPU-based model training and inference contribute significantly to carbon emissions, lightweight architectures with comparable performance to parameter-rich models present a more environmentally friendly alternative. Specifically, we build upon CycleGAN with a fixed-channel lightweight U-Net generator for modality transfer from standard confocal to super-resolution STED and deconvolved STED images, and systematically compare it against Pix2Pix and standard CycleGAN baselines. Obtaining paired datasets in medical imaging and super-resolution microscopy is often infeasible due to the need for additional experiments and the intrinsic complexity of biological sample preparation. To address this, we investigate the performance of lightweight CycleGAN models, demonstrating their ability to achieve high-fidelity modality transfer despite reduced model complexity. We introduce a fixed channel strategy within the U-Net-based generator, in contrast to the traditional channel-doubling approach. This modification significantly reduces the number of trainable parameters from 41.8 million to approximately 9 thousand, while achieving comparable or slightly improved performance. We explore the utility of GAN models as a qualitative marker for assessing experimental and labeling quality. When trained on high-quality microscopy images, the GAN implicitly learns the characteristics of optimal imaging. Deviations between GAN-generated outputs trained on high-quality data and low-quality experimental images can highlight potential issues such as photobleaching, experimental artifacts, or inaccurate labeling. In this way, the model can support qualitative assessment of experimental consistency and image fidelity in fluorescence microscopy workflows.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{articlef,

title = {High‐Speed Interferometric Scattering Tracking Microscopy of Compartmentalized Lipid Diffusion in Living Cells},

author = {Francesco Reina and Christian Eggeling and Christoffer Lagerholm},

doi = {10.1002/cphc.202400407},

year  = {2025},

date = {2025-01-01},

urldate = {2025-01-01},

journal = {ChemPhysChem},

volume = {26},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{articlei,

title = {Concurrent diffusion of nicotinic acetylcholine receptors and fluorescent cholesterol disclosed by two-colour sub-millisecond MINFLUX-based single-molecule tracking},

author = {Francesco Reina and Lucas Saavedra and Christian Eggeling and Francisco Barrantes},

doi = {10.1038/s41467-025-61489-4},

year  = {2025},

date = {2025-01-01},

urldate = {2025-01-01},

journal = {Nature Communications},

volume = {16},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{articlee,

title = {Homogeneous large field-of-view and compact iSCAT-TIRF setup for dynamic single molecule measurements},

author = {Giovanni Angelis and Jacopo Abramo and Mariia Miasnikova and Marcel Taubert and Christian Eggeling and Francesco Reina},

doi = {10.1364/OE.532947},

year  = {2024},

date = {2024-01-01},

urldate = {2024-01-01},

journal = {Optics Express},

volume = {32},

pages = {46607-46620},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{articleg,

title = {Neural network informed photon filtering reduces fluorescence correlation spectroscopy artifacts},

author = {Alexander Seltmann and Pablo Carravilla and Katharina Reglinski and Christian Eggeling and Dominic Waithe},

doi = {10.1016/j.bpj.2024.02.012},

year  = {2024},

date = {2024-01-01},

urldate = {2024-01-01},

journal = {Biophysical Journal},

volume = {123},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Svensson:23,

title = {Quantitative analysis of peroxisome tracks using a Hidden Markov Model},

author = { Carl-Magnus Svensson and Katharina Reglinski and Wolfgang Schliebs and Ralf Erdmann and Christian Eggeling and Marc Thilo Figge},

url = {https://pubmed.ncbi.nlm.nih.gov/37951993/},

doi = {10.1038/s41598-023-46812-7},

year  = {2023},

date = {2023-11-01},

urldate = {2023-11-01},

journal = {Sci Rep},

volume = {13},

number = {1},

pages = {19694},

publisher = {Springer Science and Business Media LLC},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{articlec,

title = {Characterization of diffusing sub-10 nm nano-objects using single anti-resonant element optical fibers},

author = {Torsten Wieduwilt and Ronny Förster and Mona Nissen and Jens Kobelke and Markus Schmidt},

doi = {10.1038/s41467-023-39021-3},

year  = {2023},

date = {2023-01-01},

urldate = {2023-01-01},

journal = {Nature Communications},

volume = {14},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{articleb,

title = {Nanoparticle Tracking in Single‐Antiresonant‐Element Fiber for High‐Precision Size Distribution Analysis of Mono‐ and Polydisperse Samples},

author = {Mona Nissen and Ronny Förster and Torsten Wieduwilt and Adrian Lorenz and Shiqi Jiang and Walter Hauswald and Markus Schmidt},

doi = {10.1002/smll.202202024},

year  = {2022},

date = {2022-01-01},

urldate = {2022-01-01},

journal = {Small},

volume = {18},

pages = {2202024},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{article,

title = {Ultralong Tracking of Fast diffusing Nano‐Objects Inside Nano‐Fluidic Channel Enhanced Microstructured Optical Fiber},

author = {Fengji Gui and Shiqi Jiang and Ronny Förster and Malte Plidschun and Stefan Weidlich and Jiangbo Zhao and Markus Schmidt},

doi = {10.1002/adpr.202100032},

year  = {2021},

date = {2021-01-01},

urldate = {2021-01-01},

journal = {Advanced Photonics Research},

volume = {2},

pages = {2100032},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{articled,

title = {Tracking and Analyzing the Brownian Motion of Nano-objects Inside Hollow Core Fibers},

author = {Ronny Förster and Stefan Weidlich and Mona Nissen and Torsten Wieduwilt and Jens Kobelke and Aaron Goldfain and Timothy Chiang and Rees Garmann and Vinothan Manoharan and Yoav Lahini and Markus Schmidt},

doi = {10.1021/acssensors.0c00339},

year  = {2020},

date = {2020-01-01},

urldate = {2020-01-01},

journal = {ACS Sensors},

pages = {879–886},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{doi:10.1021/acsnano.5b05646,

title = {Fast, Label-Free Tracking of Single Viruses and Weakly Scattering Nanoparticles in a Nanofluidic Optical Fiber},

author = {Sanli Faez and Yoav Lahini and Stefan Weidlich and Rees F. Garmann and Katrin Wondraczek and Matthias Zeisberger and Markus A. Schmidt and Michel Orrit and Vinothan N. Manoharan},

url = {https://doi.org/10.1021/acsnano.5b05646},

doi = {10.1021/acsnano.5b05646},

year  = {2015},

date = {2015-01-01},

urldate = {2015-01-01},

journal = {ACS Nano},

volume = {9},

number = {12},

pages = {12349-12357},

note = {PMID: 26505649},

keywords = {},

pubstate = {published},

tppubtype = {article}

}