Hybrid CNN–transformer demosaicing for bioinspired single-chip color-near-infrared fluorescence imaging in oncologic surgery
We demonstrate improvements in spatial resolution and image fidelity for color and NIR images obtained from hexachromatic imaging sensors, achieved by integrating convolutional neural networks with transformer architectures. Given recent advances in GPU computing, our CNN-transformer approach offers a practical, real-time solution for enhanced multispectral imaging during cancer surgery.
Authors
- Viktor Gruev (ORCID: https://orcid.org/0000-0003-1280-7971)
- Борислав Кондов (ORCID: https://orcid.org/0000-0002-7487-2552)
- Shaoping Nie (ORCID: https://orcid.org/0000-0002-2412-4679)
- Brian T. Cunningham (ORCID: https://orcid.org/0000-0002-9681-2104)
- Горан Кондов (ORCID: https://orcid.org/0000-0003-4567-4263)
- Yifei Jin (ORCID: https://orcid.org/0000-0002-3061-6709)
- Sunil Kumar Singhal (ORCID: https://orcid.org/0000-0002-2413-7082)
- David Forsyth (ORCID: https://orcid.org/0009-0002-1818-2052)
- Jiankun Yang (ORCID: https://orcid.org/0009-0002-2336-6606)
Institutions
- University of Illinois Urbana-Champaign (US)
- Equality Now (US)
- Ss. Cyril and Methodius University in Skopje (MK)
- University of Pennsylvania (US)
Publication Details
- Journal
- Journal of Biomedical Optics
- Published
- 2025-10-28
- DOI
- https://doi.org/10.1117/1.jbo.30.10.106008
- Primary Topic
- Advanced Optical Sensing Technologies
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- National Science Foundation
- National Institutes of Health
- U.S. Air Force