Self-supervised denoising of dynamic fluorescence images via temporal gradient-empowered deep learning

Abstract Fluorescence microscopy has become one of the most widely employed in vivo imaging modalities, enabling the discovery of new biopathological mechanisms. However, the application of fluorescence imaging is often hindered by signal-to-noise ratio issues owing to inherent noise arising from va...

Full description

Saved in:
Bibliographic Details
Main Authors: Woojin Lee, Minseok A. Jang, Hyeong Soo Nam, Jeonggeun Song, Jieun Choi, Joon Woo Song, Jae Yeon Seok, Pilhan Kim, Jin Won Kim, Hongki Yoo
Format: Article
Language:English
Published: SpringerOpen 2025-05-01
Series:PhotoniX
Online Access:https://doi.org/10.1186/s43074-025-00173-8
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849767580662759424
author Woojin Lee
Minseok A. Jang
Hyeong Soo Nam
Jeonggeun Song
Jieun Choi
Joon Woo Song
Jae Yeon Seok
Pilhan Kim
Jin Won Kim
Hongki Yoo
author_facet Woojin Lee
Minseok A. Jang
Hyeong Soo Nam
Jeonggeun Song
Jieun Choi
Joon Woo Song
Jae Yeon Seok
Pilhan Kim
Jin Won Kim
Hongki Yoo
author_sort Woojin Lee
collection DOAJ
description Abstract Fluorescence microscopy has become one of the most widely employed in vivo imaging modalities, enabling the discovery of new biopathological mechanisms. However, the application of fluorescence imaging is often hindered by signal-to-noise ratio issues owing to inherent noise arising from various systemic and biophysical characteristics. These limitations pose a growing challenge, especially with the desire to elucidate dynamic biomechanisms at previously unreachable rapid speeds. Here, we propose a temporal gradient (TG)-based self-supervised denoising network (TeD) that could enable an unprecedented advance in spatially dynamic fluorescence imaging. Our strategy is predicated on the insight that judicious utilization of spatiotemporal information is more advantageous for denoising predictions. Adopting the TG, which intrinsically embodies spatial dynamic features, enables TeD to prudently focus on spatiotemporal information. We showed that TeD can provide new interpretative opportunities for understanding dynamic fluorescence signals in in vivo imaging of mice, representing cellular flow. Furthermore, we demonstrated that TeD is robust even when fluorescence signals exhibit temporal kinetics without spatial dynamics, as seen in neuronal population imaging. We believe that TeD’s superior performance even with spatially dynamic samples, including the complex behavior of cells or organisms, could make a substantial contribution to various biological studies.
format Article
id doaj-art-adf2d0c1a7aa4ed79fb7eb6230d16ea9
institution DOAJ
issn 2662-1991
language English
publishDate 2025-05-01
publisher SpringerOpen
record_format Article
series PhotoniX
spelling doaj-art-adf2d0c1a7aa4ed79fb7eb6230d16ea92025-08-20T03:04:07ZengSpringerOpenPhotoniX2662-19912025-05-016112510.1186/s43074-025-00173-8Self-supervised denoising of dynamic fluorescence images via temporal gradient-empowered deep learningWoojin Lee0Minseok A. Jang1Hyeong Soo Nam2Jeonggeun Song3Jieun Choi4Joon Woo Song5Jae Yeon Seok6Pilhan Kim7Jin Won Kim8Hongki Yoo9Department of Mechanical Engineering, Korea Advanced Institute of Science and Technology (KAIST)Department of Mechanical Engineering, Korea Advanced Institute of Science and Technology (KAIST)Department of Mechanical Engineering, Korea Advanced Institute of Science and Technology (KAIST)Department of Mechanical Engineering, Korea Advanced Institute of Science and Technology (KAIST)Graduate School of Medical Science and Engineering, KAISTMultimodal Imaging and Theranostic Lab., Cardiovascular Center, Korea University Guro HospitalDepartment of Pathology, Yongin Severance Hospital, Yonsei University College of MedicineGraduate School of Medical Science and Engineering, KAISTMultimodal Imaging and Theranostic Lab., Cardiovascular Center, Korea University Guro HospitalDepartment of Mechanical Engineering, Korea Advanced Institute of Science and Technology (KAIST)Abstract Fluorescence microscopy has become one of the most widely employed in vivo imaging modalities, enabling the discovery of new biopathological mechanisms. However, the application of fluorescence imaging is often hindered by signal-to-noise ratio issues owing to inherent noise arising from various systemic and biophysical characteristics. These limitations pose a growing challenge, especially with the desire to elucidate dynamic biomechanisms at previously unreachable rapid speeds. Here, we propose a temporal gradient (TG)-based self-supervised denoising network (TeD) that could enable an unprecedented advance in spatially dynamic fluorescence imaging. Our strategy is predicated on the insight that judicious utilization of spatiotemporal information is more advantageous for denoising predictions. Adopting the TG, which intrinsically embodies spatial dynamic features, enables TeD to prudently focus on spatiotemporal information. We showed that TeD can provide new interpretative opportunities for understanding dynamic fluorescence signals in in vivo imaging of mice, representing cellular flow. Furthermore, we demonstrated that TeD is robust even when fluorescence signals exhibit temporal kinetics without spatial dynamics, as seen in neuronal population imaging. We believe that TeD’s superior performance even with spatially dynamic samples, including the complex behavior of cells or organisms, could make a substantial contribution to various biological studies.https://doi.org/10.1186/s43074-025-00173-8
spellingShingle Woojin Lee
Minseok A. Jang
Hyeong Soo Nam
Jeonggeun Song
Jieun Choi
Joon Woo Song
Jae Yeon Seok
Pilhan Kim
Jin Won Kim
Hongki Yoo
Self-supervised denoising of dynamic fluorescence images via temporal gradient-empowered deep learning
PhotoniX
title Self-supervised denoising of dynamic fluorescence images via temporal gradient-empowered deep learning
title_full Self-supervised denoising of dynamic fluorescence images via temporal gradient-empowered deep learning
title_fullStr Self-supervised denoising of dynamic fluorescence images via temporal gradient-empowered deep learning
title_full_unstemmed Self-supervised denoising of dynamic fluorescence images via temporal gradient-empowered deep learning
title_short Self-supervised denoising of dynamic fluorescence images via temporal gradient-empowered deep learning
title_sort self supervised denoising of dynamic fluorescence images via temporal gradient empowered deep learning
url https://doi.org/10.1186/s43074-025-00173-8
work_keys_str_mv AT woojinlee selfsuperviseddenoisingofdynamicfluorescenceimagesviatemporalgradientempowereddeeplearning
AT minseokajang selfsuperviseddenoisingofdynamicfluorescenceimagesviatemporalgradientempowereddeeplearning
AT hyeongsoonam selfsuperviseddenoisingofdynamicfluorescenceimagesviatemporalgradientempowereddeeplearning
AT jeonggeunsong selfsuperviseddenoisingofdynamicfluorescenceimagesviatemporalgradientempowereddeeplearning
AT jieunchoi selfsuperviseddenoisingofdynamicfluorescenceimagesviatemporalgradientempowereddeeplearning
AT joonwoosong selfsuperviseddenoisingofdynamicfluorescenceimagesviatemporalgradientempowereddeeplearning
AT jaeyeonseok selfsuperviseddenoisingofdynamicfluorescenceimagesviatemporalgradientempowereddeeplearning
AT pilhankim selfsuperviseddenoisingofdynamicfluorescenceimagesviatemporalgradientempowereddeeplearning
AT jinwonkim selfsuperviseddenoisingofdynamicfluorescenceimagesviatemporalgradientempowereddeeplearning
AT hongkiyoo selfsuperviseddenoisingofdynamicfluorescenceimagesviatemporalgradientempowereddeeplearning