# Deep Learning-Enabled Virtual Multiplexed Immunostaining of Label-Free Tissue for Vascular Invasion Assessment

## TL;DR

This paper introduces a deep learning method to create virtual immunostaining images from label-free tissue, improving the assessment of vascular invasion in thyroid cancer without traditional staining.

## Contribution

A novel deep learning framework for virtual multiplexed immunostaining that eliminates the need for chemical staining and enables accurate vascular invasion assessment.

## Key findings

- Virtual mIHC staining showed high concordance with traditional histochemical staining results.
- The method accurately localized epithelial and endothelial cells and identified small vessel invasion.
- The approach reduces tissue loss and variability compared to conventional IHC techniques.

## Abstract

Objective: We report the development and validation of a deep learning-based virtual multiplexed immunostaining method for label-free tissue, enabling the simultaneous generation of ERG (ETS-related gene), PanCK (pan-cytokeratin), and hematoxylin and eosin (H&E) images for vascular invasion assessment. Impact Statement: This work delivers routine laboratory-compatible virtual multiplexed immunohistochemistry (mIHC) that reproduces ERG, PanCK, and H&E on the same tissue section without chemical staining. It addresses the cost, labor, tissue loss, and section-to-section variability of conventional IHC, as well as the practical unavailability of mIHC in most pathology laboratories, thereby improving accuracy and efficiency in assessing vascular invasion. Introduction: Traditional IHC requires one tissue section per stain, exhibits section-to-section variability, and incurs high costs and laborious staining procedures. While 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 that simultaneously generates ERG and PanCK, in addition to H&E virtual staining, enabling the accurate localization and interpretation of vascular invasion in thyroid cancers. Methods: 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. Results: Blind evaluation by board-certified pathologists demonstrated that virtual mIHC staining achieved high concordance with the histochemical staining results, accurately highlighting epithelial and endothelial cells. Virtual mIHC conducted on the same tissue section also allowed the identification and localization of small vessel invasion. Conclusion: This virtual mIHC approach can substantially 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.