MOSBY enables multi-omic inference and spatial biomarker discovery from whole slide images DOI Creative Commons
Yasin Şenbabaoğlu, Vignesh Prabhakar, Aminollah Khormali

et al.

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Aug. 6, 2024

The utility of deep neural nets has been demonstrated for mapping hematoxylin-and-eosin (H&E) stained image features to expression individual genes. However, these models have not employed discover clinically relevant spatial biomarkers. Here we develop MOSBY (Multi-Omic translation whole slide images Spatial Biomarker discoverY) that leverages contrastive self-supervised pretraining extract improved H&E features, learns a between and bulk omic profiles (RNA, DNA, protein), utilizes tile-level information We validate gene set predictions with transcriptomic serially-sectioned CD8 IHC data. demonstrate MOSBY-inferred colocalization survival-predictive power orthogonal expression, enable concordance indices highly competitive survival-trained multimodal networks. identify (1) an ER stress-associated feature as chemotherapy-specific risk factor in lung adenocarcinoma, (2) the T effector cell vs cysteine signatures negative prognostic multiple cancer indications. discovery biologically interpretable biomarkers showcases model unraveling novel insights biology well informing clinical decision-making.

Language: Английский

A deep-learning framework to predict cancer treatment response from histopathology images through imputed transcriptomics DOI
Danh-Tai Hoang, Gal Dinstag, Eldad D. Shulman

et al.

Nature Cancer, Journal Year: 2024, Volume and Issue: 5(9), P. 1305 - 1317

Published: July 3, 2024

Language: Английский

Citations

29
Self-interactive learning: Fusion and evolution of multi-scale histomorphology features for molecular traits prediction in computational pathology DOI Creative Commons
Yang Hu, Korsuk Sirinukunwattana, Bin Li

et al.

Medical Image Analysis, Journal Year: 2025, Volume and Issue: 101, P. 103437 - 103437

Published: Jan. 5, 2025

Language: Английский

Citations

2
Revolutionizing Digital Pathology With the Power of Generative Artificial Intelligence and Foundation Models DOI Creative Commons
Asim Waqas, Marilyn M. Bui,

Eric F. Glassy

et al.

Laboratory Investigation, Journal Year: 2023, Volume and Issue: 103(11), P. 100255 - 100255

Published: Sept. 26, 2023

Language: Английский

Citations

38
A survey of Transformer applications for histopathological image analysis: New developments and future directions DOI Creative Commons
Chukwuemeka Clinton Atabansi, Jing Nie, Haijun Liu

et al.

BioMedical Engineering OnLine, Journal Year: 2023, Volume and Issue: 22(1)

Published: Sept. 25, 2023

Transformers have been widely used in many computer vision challenges and shown the capability of producing better results than convolutional neural networks (CNNs). Taking advantage capturing long-range contextual information learning more complex relations image data, applied to histopathological processing tasks. In this survey, we make an effort present a thorough analysis uses analysis, covering several topics, from newly built Transformer models unresolved challenges. To be precise, first begin by outlining fundamental principles attention mechanism included other key frameworks. Second, analyze Transformer-based applications imaging domain provide evaluation 100 research publications across different downstream tasks cover most recent innovations, including survival prediction, segmentation, classification, detection, representation. Within survey work, also compare performance CNN-based techniques based on recently published papers, highlight major challenges, interesting future directions. Despite outstanding architectures number papers reviewed anticipate that further improvements exploration are still required future. We hope paper will give readers field study understanding up-to-date list summary provided at https://github.com/S-domain/Survey-Paper .

Language: Английский

Citations

26
Prediction of DNA methylation-based tumor types from histopathology in central nervous system tumors with deep learning DOI
Danh-Tai Hoang, Eldad D. Shulman,

Rust Turakulov

et al.

Nature Medicine, Journal Year: 2024, Volume and Issue: 30(7), P. 1952 - 1961

Published: May 17, 2024

Language: Английский

Citations

15
Histopathology based AI model predicts anti-angiogenic therapy response in renal cancer clinical trial DOI Creative Commons

Jay Jasti,

Hua Zhong,

Vandana Panwar

et al.

Nature Communications, Journal Year: 2025, Volume and Issue: 16(1)

Published: March 17, 2025

Abstract Anti-angiogenic (AA) therapy is a cornerstone of metastatic clear cell renal carcinoma (ccRCC) treatment, but not everyone responds, and predictive biomarkers are lacking. CD31, marker vasculature, insufficient, the Angioscore, an RNA-based angiogenesis quantification method, costly, associated with delays, difficult to standardize, does account for tumor heterogeneity. Here, we developed interpretable deep learning (DL) model that predicts Angioscore directly from ubiquitous histopathology slides yielding visual vascular network (H&E DL Angio). H&E Angio achieves strong correlation across multiple cohorts (spearman correlations 0.77 0.73). Using this approach, found inversely correlates grade stage driver mutation status. Importantly, expediently AA response in both real-world IMmotion150 trial cohorts, out-performing closely approximating (c-index 0.66 vs 0.67) at fraction cost.

Language: Английский

Citations

1
On image search in histopathology DOI Creative Commons
Hamid R. Tizhoosh, Liron Pantanowitz

Journal of Pathology Informatics, Journal Year: 2024, Volume and Issue: 15, P. 100375 - 100375

Published: April 5, 2024

Pathology images of histopathology can be acquired from camera-mounted microscopes or whole-slide scanners. Utilizing similarity calculations to match patients based on these holds significant potential in research and clinical contexts. Recent advancements search technologies allow for implicit quantification tissue morphology across diverse primary sites, facilitating comparisons, enabling inferences about diagnosis, potentially prognosis, predictions new when compared against a curated database diagnosed treated cases. In this article, we comprehensively review the latest developments image histopathology, offering concise overview tailored computational pathology researchers seeking effective, fast, efficient methods their work.

Language: Английский

Citations

5
One label is all you need: Interpretable AI-enhanced histopathology for oncology DOI
Thomas E. Tavolara, Ziyu Su, Metin N. Gürcan

et al.

Seminars in Cancer Biology, Journal Year: 2023, Volume and Issue: 97, P. 70 - 85

Published: Oct. 11, 2023

Language: Английский

Citations

11
Digital profiling of gene expression from histology images with linearized attention DOI Creative Commons
Marija Pizurica, Yuanning Zheng, Francisco Carrillo‐Pérez

et al.

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: Nov. 14, 2024

Cancer is a heterogeneous disease requiring costly genetic profiling for better understanding and management. Recent advances in deep learning have enabled cost-effective predictions of alterations from whole slide images (WSIs). While transformers driven significant progress non-medical domains, their application to WSIs lags behind due high model complexity limited dataset sizes. Here, we introduce SEQUOIA, linearized transformer that predicts cancer transcriptomic profiles WSIs. SEQUOIA developed using 7584 tumor samples across 16 types, with its generalization capacity validated on two independent cohorts comprising 1368 tumors. Accurately predicted genes are associated key processes, including inflammatory response, cell cycles metabolism. Further, demonstrate the value stratifying risk breast recurrence resolving spatial gene expression at loco-regional levels. hence deciphers clinically relevant information WSIs, opening avenues personalized Predicting whole-slide (WSIs) can be cost-efficient solution profiling. authors develop linearised attention predict validate performance clinical utility multiple pan-cancer datasets.

Language: Английский

Citations

4
Efficient merging and validation of deep learning-based nuclei segmentations in H&E slides from multiple models DOI Creative Commons
Jagadheshwar Balan, Shannon K. McDonnell, Zachary C. Fogarty

et al.

Journal of Pathology Informatics, Journal Year: 2025, Volume and Issue: unknown, P. 100443 - 100443

Published: April 1, 2025

Language: Английский

Citations

0