Pan-cancer image-based detection of clinically actionable genetic alterations
Пан-канцерная детекция клинически значимых генетических изменений по изображениям
2020-07-27
SCID: 54.1/zgfyz9tm
Discuss with AI
H&E-stained paraffin-embedded tissuedeep learning for genetic alteration predictionhistopathology whole-slide imagesimage-based detectionpan-cancer
Figures from the paper
Abstract (AI)
Molecular alterations in cancer can cause phenotypic changes in tumor cells and their micro-environment. Routine histopathology tissue slides - which are ubiquitously available - can reflect such morphological changes. Here, we show that deep learning can consistently infer a wide range of genetic mutations, molecular tumor subtypes, gene expression signatures and standard pathology biomarkers directly from routine histology. We developed, optimized, validated and publicly released a one-stop-shop workflow and applied it to tissue slides of more than 5000 patients across multiple solid tumors. Our findings show that a single deep learning algorithm can be trained to predict a wide range of molecular alterations from routine, paraffin-embedded histology slides stained with hematoxylin and eosin. These predictions generalize to other populations and are spatially resolved. Our method can be implemented on mobile hardware, potentially enabling point-of-care diagnostics for personalized cancer treatment. More generally, this approach could elucidate and quantify genotype-phenotype links in cancer.
Key Findings
1
A single deep learning algorithm can infer a wide range of genetic mutations, molecular tumor subtypes, gene expression signatures, and standard pathology biomarkers from routine H&E-stained histology slides.
2
Predictions from the model generalize to other populations and are spatially resolved within histological images.
3
The authors developed, optimized, validated, and publicly released a one-stop-shop workflow applied to tissue slides from over 5,000 patients across multiple solid tumor types.
4
The method can be implemented on mobile hardware, enabling potential point-of-care diagnostics for personalized cancer treatment.
5
This image-based approach can elucidate and quantify genotype–phenotype links in cancer by linking molecular alterations to morphological patterns.
Research Object
Routine hematoxylin-and-eosin stained paraffin-embedded histology slides of multiple solid tumors (pan-cancer tissue slides)
Research Subject
Deep-learning-based inference and spatially resolved prediction of clinically actionable molecular alterations (genetic mutations, molecular subtypes, gene expression signatures, and pathology biomarkers) from routine histology, including generalizability and point-of-care implementation
Publication Details
Publication Date
2020-07-27
Journal
Publisher
ISSN
Cited by
670
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest
References available in scid.ai3
Cited by4
Deep learning in cancer diagnosis, prognosis and treatment selection2021
Pan-cancer integrative histology-genomic analysis via multimodal deep learning2022
AI-enabled virtual spatial proteomics from histopathology for interpretable biomarker discovery in lung cancer2026
Will AI Replace Physicians in the Near Future? AI Adoption Barriers in Medicine2026