A computational super-resolution framework for multidimensional fluorescence imaging
8 affiliations
- 1Wellman Center for Photomedicine, Harvard Medical School, Massachusetts General Hospital, CNY149, 13th St, Charlestown, MA 02129, USA.
- 2Department of Electrical and Computer Engineering, Boston University, 8 St Mary’s St, Boston, MA 02215, USA.
- 3College of Graduate and Professional Studies, Trine University, 999 Republic Dr Suite 200, Allen Park, Detroit, MI 48101, USA.
- 4Laboratory for Optical and Computational Instrumentation, University of Wisconsin-Madison, Madison, WI 53706, USA.
- 5Intek Scientific, 1 Broadway, Cambridge, MA 02142, USA.
- 6Laboratorio Nacional de Microscopía Avanzada, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Cuernavaca, Morelos, Mexico.
- 7Department of Biomedical Engineering, Boston University, 44 Cummington Mall, Boston, MA 02215, USA.
- 8Center for Interdisciplinary Innovation in Imaging, Massachusetts General Hospital and Harvard Medical School, 149 13th St, Charlestown, MA 02129, USA.
Abstract
Fully capturing the heterogeneity of biological processes remains a central challenge, as conventional confocal microscopy typically surveys large ensembles of molecules within a small volume. The development of super-resolution imaging enables the capture of images with exceptional detail. When super-resolution is combined with functional imaging, such as fluorescence lifetime and spectral imaging, it provides better separation of image constituents and sensing of environmental properties at the nanoscale. These multidimensional imaging approaches often acquire functional and fluorescence images separately and then merge the resulting datasets, a post hoc approach that risks dynamic range mismatches and the loss of subtle molecular signals. Here, we introduce a computational super-resolution framework based on the deblurring by pixel reassignment (DPR) algorithm, which processes multidimensional data as high-dimensional tensors, integrating spatial axes with fluorescence lifetime or spectral information into a single dataset. Using DPR, high-resolution details can be extracted from conventional fluorescence microscopy images, even from a single capture. We show that this approach enhances the spatial resolution of multidimensional datasets without compromising quantitative fidelity. As a proof of concept, we demonstrate super-resolution multimodal imaging (intensity, lifetime, and spectra), revealing fine-scale details while preserving signals from rare or low-intensity molecular events. This accessible yet powerful method paves the way for quantitative, single-molecule-level insights into complex biological systems.
Related
- Low-affinity exchangeable probes for super-resolution and long-term fluorescence imaging.
- Vision Transformers as a Framework for Imaging Fluorescence Correlation Spectroscopy
- Fluorescence imaging-induced phospholipid oxidation drives membrane phase separation by shifting miscibility boundaries
- Application of a light emitting diode based autofluorescence removal system to improve fluorescence imaging in plant tissues
- Quantitative fluorescence lifetime imaging of protein microenvironments by genetically encoded compact benzodiazole probes
BibTeX
@article{hwang2026computational,
title = {{A computational super-resolution framework for multidimensional fluorescence imaging}},
author = {Wonsang Hwang and Bingying Zhao and Kai Guo and Jenu V. Chacko and Sinyoung Jeong and Adán Guerrero and Jerome Mertz and Conor L. Evans and Iván Coto Hernández},
journal = {Science Advances},
year = {2026},
month = oct,
doi = {10.1126/sciadv.aec7970},
url = {https://www.science.org/doi/10.1126/sciadv.aec7970},
note = {CC BY-NC},
}
RIS
TY - JOUR
TI - A computational super-resolution framework for multidimensional fluorescence imaging
AU - Wonsang Hwang
AU - Bingying Zhao
AU - Kai Guo
AU - Jenu V. Chacko
AU - Sinyoung Jeong
AU - Adán Guerrero
AU - Jerome Mertz
AU - Conor L. Evans
AU - Iván Coto Hernández
PY - 2026
DA - 2026/10/02
JO - Science Advances
DO - 10.1126/sciadv.aec7970
UR - https://www.science.org/doi/10.1126/sciadv.aec7970
AB - Fully capturing the heterogeneity of biological processes remains a central challenge, as conventional confocal microscopy typically surveys large ensembles of molecules within a small volume. The development of super-resolution imaging enables the capture of images with exceptional detail. When super-resolution is combined with functional imaging, such as fluorescence lifetime and spectral imaging, it provides better separation of image constituents and sensing of environmental properties at the nanoscale. These multidimensional imaging approaches often acquire functional and fluorescence images separately and then merge the resulting datasets, a post hoc approach that risks dynamic range mismatches and the loss of subtle molecular signals. Here, we introduce a computational super-resolution framework based on the deblurring by pixel reassignment (DPR) algorithm, which processes multidimensional data as high-dimensional tensors, integrating spatial axes with fluorescence lifetime or spectral information into a single dataset. Using DPR, high-resolution details can be extracted from conventional fluorescence microscopy images, even from a single capture. We show that this approach enhances the spatial resolution of multidimensional datasets without compromising quantitative fidelity. As a proof of concept, we demonstrate super-resolution multimodal imaging (intensity, lifetime, and spectra), revealing fine-scale details while preserving signals from rare or low-intensity molecular events. This accessible yet powerful method paves the way for quantitative, single-molecule-level insights into complex biological systems.
SN - 2375-2548
ER -