Deep learning-enabled virtual multiplexed immunostaining of label-free tissue for vascular invasion assessment

📅 2025-08-22
📈 Citations: 0
✨ Influential: 0
📄 PDF
🤖 AI Summary
Conventional immunohistochemistry (IHC) and multiplex IHC (mIHC) rely on physical staining of tissue sections, leading to tissue consumption, inter-slide variability, high cost, technical complexity, and limited clinical deployability. To address these limitations, we propose a label-free virtual multiplex staining framework based on autofluorescence imaging. Leveraging deep learning, our method directly synthesizes virtual ERG, PanCK, and H&E-stained images from a single unstained thyroid tissue section—enabling simultaneous multi-marker analysis on the same slide for the first time. By eliminating the entire physical staining workflow, the approach significantly reduces tissue usage and operational barriers. Quantitative evaluation shows high fidelity between virtual and ground-truth stains (PSNR > 30 dB), and blinded pathological assessment confirms accurate epithelial and endothelial cell identification, as well as reliable detection of microvascular invasion. This paradigm offers a scalable, cost-effective solution for precision diagnosis of thyroid carcinoma.

Technology Category

Application Category

📝 Abstract
Immunohistochemistry (IHC) has transformed clinical pathology by enabling the visualization of specific proteins within tissue sections. However, traditional IHC requires one tissue section per stain, exhibits section-to-section variability, and incurs high costs and laborious staining procedures. While multiplexed IHC (mIHC) techniques enable simultaneous staining with multiple antibodies on a single slide, they are more tedious to perform and are currently unavailable in routine pathology laboratories. Here, we present a deep learning-based virtual multiplexed immunostaining framework to simultaneously generate ERG and PanCK, in addition to H&E virtual staining, enabling accurate localization and interpretation of vascular invasion in thyroid cancers. This virtual mIHC technique is based on the autofluorescence microscopy images of label-free tissue sections, and its output images closely match the histochemical staining counterparts (ERG, PanCK and H&E) of the same tissue sections. Blind evaluation by board-certified pathologists demonstrated that virtual mIHC staining achieved high concordance with the histochemical staining results, accurately highlighting epithelial cells and endothelial cells. Virtual mIHC conducted on the same tissue section also allowed the identification and localization of small vessel invasion. This multiplexed virtual IHC approach can significantly improve diagnostic accuracy and efficiency in the histopathological evaluation of vascular invasion, potentially eliminating the need for traditional staining protocols and mitigating issues related to tissue loss and heterogeneity.
Problem

Research questions and friction points this paper is trying to address.

Virtual multiplexed immunostaining replaces traditional IHC staining
Deep learning generates multiple stains from label-free tissue
Enables accurate vascular invasion assessment in thyroid cancers
Innovation

Methods, ideas, or system contributions that make the work stand out.

Deep learning generates virtual multiplexed immunostaining from label-free tissue
Uses autofluorescence microscopy images to create ERG, PanCK, H&E stains
Enables accurate vascular invasion localization without traditional staining
🔎 Similar Papers
No similar papers found.
Authors to Follow
Y
Yijie Zhang
Electrical and Computer Engineering Department, University of California, Los Angeles, CA, 90095, USA; Bioengineering Department, University of California, Los Angeles, CA, 90095, USA; California NanoSystems Institute (CNSI), University of California, Los Angeles, CA, 90095, USA
C
Cagatay Isil
Electrical and Computer Engineering Department, University of California, Los Angeles, CA, 90095, USA; Bioengineering Department, University of California, Los Angeles, CA, 90095, USA; California NanoSystems Institute (CNSI), University of California, Los Angeles, CA, 90095, USA
X
Xilin Yang
Electrical and Computer Engineering Department, University of California, Los Angeles, CA, 90095, USA; Bioengineering Department, University of California, Los Angeles, CA, 90095, USA; California NanoSystems Institute (CNSI), University of California, Los Angeles, CA, 90095, USA
Yuzhu Li
Yuzhu Li
University of California, Los Angeles
Computational imagingOptical imaging and sensingMachine learning
A
Anna Elia
Department of Pathology, Hadassah Hebrew University Medical Center, Jerusalem, 91120, Israel
K
Karin Atlan
Department of Pathology, Hadassah Hebrew University Medical Center, Jerusalem, 91120, Israel
W
William Dean Wallace
Department of Pathology, Keck School of Medicine, University of Southern California, Los Angeles, CA, 90033, USA
N
Nir Pillar
Electrical and Computer Engineering Department, University of California, Los Angeles, CA, 90095, USA; Bioengineering Department, University of California, Los Angeles, CA, 90095, USA; California NanoSystems Institute (CNSI), University of California, Los Angeles, CA, 90095, USA; Department of Pathology, Hadassah Hebrew University Medical Center, Jerusalem, 91120, Israel
Aydogan Ozcan
Aydogan Ozcan
Chancellor's Professor at UCLA & HHMI Professor
Computational ImagingHolographyMicroscopySensingBioPhotonics