Abstract:Multiplexed immunofluorescence (IF) staining plays a crucial role in cancer research, allowing simultaneous visualization of multiple biomarkers to elucidate complex cellular interactions within the tumor microenvironment. However, the conventional multiplexed IF approach has inherent limitations, characterized by labor-intensiveness and a constrained capacity for high-level multiplexing. These challenges hinder its scalability and comprehensive analysis of intricate molecular landscapes.To overcome the limitations of conventional multiplexed IF, we propose a novel approach that integrates holotomography (HT) and deep learning. HT is a label-free imaging technique that quantitatively measures the refractive index distribution of the specimen without any staining. To provide subcellular specificity to the label-free HT images, HT is integrated with virtual staining. Virtual staining is a deep learning-based framework that converts label-free images to the desired stained images, in this case, from HT images to multiplexed IF images. This synergistic combination of HT and deep learning may address the current limitations of multiplexed IF protocols.Our proposed methodology aims at high-level multiplexed virtual IF staining from label-free cancer tissue slides using HT and deep learning. This approach not only eliminates the need for chemical staining but also enhances the level of multiplexing through computational staining. Specifically, we acquire HT and fluorescence images of stained tissue slides for training. Utilizing this paired dataset, we then employ supervised learning to create a neural network that can map between label-free HT images and corresponding fluorescence images. Subsequently, the trained neural network is applied to label-free tissue slides, inferring fluorescence images without the need for chemical staining.We believe the proposed virtual staining framework will contribute to a more efficient and scalable paradigm for cancer research, facilitating a deeper understanding of tumor biology, and advancing the field of precision medicine. Citation Format: Juyeon Park, Su-Jin Shin, Geon Kim, Kwang Suk Lee, Ji Eun Heo, YongKeun Park. High-level multiplexed virtual immunofluorescence staining using holotomography and deep learning [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular s); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl) nr 7401.

Fluoroformer: Scaling multiple instance learning to multiplexed images via attention-based channel fusion

Towards Efficient Information Fusion: Concentric Dual Fusion Attention Based Multiple Instance Learning for Whole Slide Images

Finding Regions of Interest in Whole Slide Images Using Multiple Instance Learning

MAMILNet: advancing precision oncology with multi-scale attentional multi-instance learning for whole slide image analysis

Fluorescence Lifetime Multiplexing (FLEX) for simultaneous high dimensional spatial biology in 3D

DTFD-MIL: Double-Tier Feature Distillation Multiple Instance Learning for Histopathology Whole Slide Image Classification

FR-MIL: Distribution Re-calibration based Multiple Instance Learning with Transformer for Whole Slide Image Classification

Iterative multiple instance learning for weakly annotated whole slide image classification

PreMix: Boosting Multiple Instance Learning in Digital Histopathology through Pre-training with Intra-Batch Slide Mixing

FourierMIL: Fourier filtering-based multiple instance learning for whole slide image analysis

Abstract 7401: High-level multiplexed virtual immunofluorescence staining using holotomography and deep learning

RoFormer for Position Aware Multiple Instance Learning in Whole Slide Image Classification

Let it shine: Autofluorescence of Papanicolaou-stain improves AI-based cytological oral cancer detection

Deep Learning-Inferred Multiplex ImmunoFluorescence for IHC Image Quantification

Advances in Multiple Instance Learning for Whole Slide Image Analysis: Techniques, Challenges, and Future Directions

Dual-Attention Multiple Instance Learning Framework for Pathology Whole-Slide Image Classification

Long-MIL: Scaling Long Contextual Multiple Instance Learning for Histopathology Whole Slide Image Analysis

Multimodal Outer Arithmetic Block Dual Fusion of Whole Slide Images and Omics Data for Precision Oncology

Single-detector multiplex imaging flow cytometry for cancer cell classification with deep learning

Shared-Specific Feature Learning with Bottleneck Fusion Transformer for Multi-Modal Whole Slide Image Analysis.

Spatial Immunophenotyping from Whole-Slide Multiplexed Tissue Imaging Using Convolutional Neural Networks