Rapid cancer diagnosis using deep learning–powered label-free subcellular-resolution photoacoustic histology
Open Access
- 21 November 2025
- journal article
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 11 (47) , eadz1820
Abstract
Traditional hematoxylin and eosin staining in formalin-fixed paraffin-embedded sections, while essential for diagnostic pathology, is time-consuming, labor intensive, and prone to artifacts that can obscure critical histological details. Label-free ultraviolet photoacoustic microscopy (UV-PAM) has emerged as a promising alternative, offering fast histology-like images without the need for traditional staining and excessive tissue preparation. However, current UV-PAM systems face challenges in achieving the high spatial resolution required for detailed histological analysis and diagnosis. To address this, we developed a subcellular-resolution UV-PAM (SRUV-PAM) system with a 240-nanometer resolution, enabled by the integration of a high numerical aperture (NA) objective lens (NA = 0.64) and the precise piezo actuators for fine scanning control. This configuration allows visualization of detailed nuclear structures. In addition, we demonstrated virtual staining of SRUV-PAM images via cycle-consistent generative adversarial networks and diagnosis of malignant and benign tumors in liver tissues via densely connected convolutional networks DenseNet-121, achieving an area under the receiver operating characteristic curve of 0.902.Keywords
Funding Information
- National Institutes of Health (R01 EB028277, R35 CA220436)
- National Institutes of Health (R00 EB034298, T32 EB027629)
- National Institutes of Health (NIGMS T32 GM008042, NIGMS T32 GM152342)
- National Institutes of Health (K99 EB035645, T32 HL144449)
- Ministry of Education (BK21 FOUR Project)
- National Research Foundation of Korea (RS-2023-00266110, RS-2023-00210682, 2020R1A6A1A03047902, RS-2024-00462912)
This publication has 36 references indexed in Scilit:
- Large scale tissue histopathology image classification, segmentation, and visualization via deep convolutional activation featuresBMC Bioinformatics, 2017
- A practical guide to photoacoustic tomography in the life sciencesNature Methods, 2016
- Tissue Processing and Hematoxylin and Eosin Staining2014
- Optimization and enhancement of H&E stained microscopical images by applying bilinear interpolation method on lab color modeTheoretical Biology and Medical Modelling, 2014
- PHOTOACOUSTIC TOMOGRAPHY: PRINCIPLES AND ADVANCES (Invited Review)Progress In Electromagnetics Research, 2014
- In vivo label-free photoacoustic microscopy of cell nuclei by excitation of DNA and RNAOptics Letters, 2010
- Intraoperative Evaluation of Breast Tumor Margins with Optical Coherence TomographyCancer Research, 2009
- Multiscale photoacoustic microscopy and computed tomographyNature Photonics, 2009
- Hematoxylin and Eosin Staining of Tissue and Cell SectionsCold Spring Harbor Protocols, 2008
- Revised UV extinction coefficients for nucleoside-5′-monophosphates and unpaired DNA and RNANucleic Acids Research, 2004