Artificial intelligence in histopathology: enhancing cancer research and clinical oncology

Nature Cancer, Journal Year: 2022, Volume and Issue: 3(9), P. 1026 - 1038

Published: Sept. 22, 2022

Language: Английский

---

Swarm learning for decentralized artificial intelligence in cancer histopathology

Nature Medicine, Journal Year: 2022, Volume and Issue: 28(6), P. 1232 - 1239

Published: April 25, 2022

Abstract Artificial intelligence (AI) can predict the presence of molecular alterations directly from routine histopathology slides. However, training robust AI systems requires large datasets for which data collection faces practical, ethical and legal obstacles. These obstacles could be overcome with swarm learning (SL), in partners jointly train models while avoiding transfer monopolistic governance. Here, we demonstrate successful use SL large, multicentric gigapixel images over 5,000 patients. We show that trained using BRAF mutational status microsatellite instability hematoxylin eosin (H&E)-stained pathology slides colorectal cancer. on three patient cohorts Northern Ireland, Germany United States, validated prediction performance two independent Kingdom. Our SL-trained outperform most locally models, perform par are merged datasets. In addition, SL-based efficient. future, used to distributed any image analysis task, eliminating need transfer.

Language: Английский

---

Benchmarking weakly-supervised deep learning pipelines for whole slide classification in computational pathology

Medical Image Analysis, Journal Year: 2022, Volume and Issue: 79, P. 102474 - 102474

Published: May 5, 2022

Language: Английский

---

Tumour-infiltrating lymphocytes: from prognosis to treatment selection

British Journal of Cancer, Journal Year: 2022, Volume and Issue: 128(3), P. 451 - 458

Published: Dec. 23, 2022

Abstract Tumour-infiltrating lymphocytes (TILs) are considered crucial in anti-tumour immunity. Accordingly, the presence of TILs contains prognostic and predictive value. In 2011, we performed a systematic review meta-analysis on value across cancer types. Since then, advent immune checkpoint blockade (ICB) has renewed interest analysis TILs. this review, first describe how our understanding TIL changed over last decade. New insights novel subsets discussed give broader view effect cancer. Apart from value, evidence significance therapy era discussed, as well new techniques, such machine learning that strive to incorporate these capacities within clinical trials.

Language: Английский

---

Transformer-based biomarker prediction from colorectal cancer histology: A large-scale multicentric study

Cancer Cell, Journal Year: 2023, Volume and Issue: 41(9), P. 1650 - 1661.e4

Published: Aug. 30, 2023

Deep learning (DL) can accelerate the prediction of prognostic biomarkers from routine pathology slides in colorectal cancer (CRC). However, current approaches rely on convolutional neural networks (CNNs) and have mostly been validated small patient cohorts. Here, we develop a new transformer-based pipeline for end-to-end biomarker by combining pre-trained transformer encoder with network patch aggregation. Our approach substantially improves performance, generalizability, data efficiency, interpretability as compared state-of-the-art algorithms. After training evaluating large multicenter cohort over 13,000 patients 16 cohorts, achieve sensitivity 0.99 negative predictive value microsatellite instability (MSI) surgical resection specimens. We demonstrate that specimen-only reaches clinical-grade performance endoscopic biopsy tissue, solving long-standing diagnostic problem.

Language: Английский

---

Artificial intelligence to identify genetic alterations in conventional histopathology

The Journal of Pathology, Journal Year: 2022, Volume and Issue: 257(4), P. 430 - 444

Published: March 28, 2022

Precision oncology relies on the identification of targetable molecular alterations in tumor tissues. In many types, a limited set tests is currently part standard diagnostic workflows. However, universal testing for all alterations, especially rare ones, by cost and availability assays. From 2017 to 2021, multiple studies have shown that artificial intelligence (AI) methods can predict probability specific genetic directly from conventional hematoxylin eosin (H&E) tissue slides. Although these are less accurate than gold (e.g. immunohistochemistry, polymerase chain reaction or next-generation sequencing), they could be used as pre-screening tools reduce workload analyses. this systematic literature review, we summarize state art predicting H&E using AI. We found AI perform reasonably well across although few algorithms been broadly validated. addition, FGFR, IDH, PIK3CA, BRAF, TP53, DNA repair pathways predictable while other rarely investigated were only poorly predictable. Finally, discuss next steps implementation AI-based surrogate © 2022 The Authors. Journal Pathology published John Wiley & Sons Ltd behalf Pathological Society Great Britain Ireland.

Language: Английский

---

Deep Learning on Histopathological Images for Colorectal Cancer Diagnosis: A Systematic Review

Diagnostics, Journal Year: 2022, Volume and Issue: 12(4), P. 837 - 837

Published: March 29, 2022

Colorectal cancer (CRC) is the second most common in women and third men, with an increasing incidence. Pathology diagnosis complemented prognostic predictive biomarker information first step for personalized treatment. The increased diagnostic load pathology laboratory, combined reported intra- inter-variability assessment of biomarkers, has prompted quest reliable machine-based methods to be incorporated into routine practice. Recently, Artificial Intelligence (AI) made significant progress medical field, showing potential clinical applications. Herein, we aim systematically review current research on AI CRC image analysis. In histopathology, algorithms based Deep Learning (DL) have assist diagnosis, predict clinically relevant molecular phenotypes microsatellite instability, identify histological features related prognosis correlated metastasis, assess specific components tumor microenvironment.

Language: Английский

---

Generalizable biomarker prediction from cancer pathology slides with self-supervised deep learning: A retrospective multi-centric study

Cell Reports Medicine, Journal Year: 2023, Volume and Issue: 4(4), P. 100980 - 100980

Published: March 22, 2023

Deep learning (DL) can predict microsatellite instability (MSI) from routine histopathology slides of colorectal cancer (CRC). However, it is unclear whether DL also other biomarkers with high performance and predictions generalize to external patient populations. Here, we acquire CRC tissue samples two large multi-centric studies. We systematically compare six different state-of-the-art architectures pathology slides, including MSI mutations in BRAF, KRAS, NRAS, PIK3CA. Using a validation cohort provide realistic evaluation setting, show that models using self-supervised, attention-based multiple-instance consistently outperform previous approaches while offering explainable visualizations the indicative regions morphologies. While prediction BRAF reaches clinical-grade performance, mutation PIK3CA, NRAS was clinically insufficient.

Language: Английский

---

Computational Pathology: A Survey Review and The Way Forward

Journal of Pathology Informatics, Journal Year: 2024, Volume and Issue: unknown, P. 100357 - 100357

Published: Jan. 1, 2024

Computational Pathology (CPath) is an interdisciplinary science that augments developments of computational approaches to analyze and model medical histopathology images. The main objective for CPath develop infrastructure workflows digital diagnostics as assistive CAD system clinical pathology, facilitating transformational changes in the diagnosis treatment cancer are mainly address by tools. With evergrowing deep learning computer vision algorithms, ease data flow from currently witnessing a paradigm shift. Despite sheer volume engineering scientific works being introduced image analysis, there still considerable gap adopting integrating these algorithms practice. This raises significant question regarding direction trends undertaken CPath. In this article we provide comprehensive review more than 800 papers challenges faced problem design all-the-way application implementation viewpoints. We have catalogued each paper into model-card examining key layout current landscape hope helps community locate relevant facilitate understanding field's future directions. nutshell, oversee cycle stages which required be cohesively linked together associated with such multidisciplinary science. overview different perspectives data-centric, model-centric, application-centric problems. finally sketch remaining directions technical integration For updated information on survey accessing original cards repository, please refer GitHub. Updated version draft can also found arXiv.

Language: Английский

---

Artificial intelligence in digital pathology: a systematic review and meta-analysis of diagnostic test accuracy

npj Digital Medicine, Journal Year: 2024, Volume and Issue: 7(1)

Published: May 4, 2024

Abstract Ensuring diagnostic performance of artificial intelligence (AI) before introduction into clinical practice is essential. Growing numbers studies using AI for digital pathology have been reported over recent years. The aim this work to examine the accuracy in images any disease. This systematic review and meta-analysis included type applied whole slide (WSIs) reference standard was diagnosis by histopathological assessment and/or immunohistochemistry. Searches were conducted PubMed, EMBASE CENTRAL June 2022. Risk bias concerns applicability assessed QUADAS-2 tool. Data extraction two investigators performed a bivariate random effects model, with additional subgroup analyses also performed. Of 2976 identified studies, 100 48 meta-analysis. Studies from range countries, including 152,000 (WSIs), representing many diseases. These mean sensitivity 96.3% (CI 94.1–97.7) specificity 93.3% 90.5–95.4). There heterogeneity study design 99% inclusion had at least one area high or unclear risk concerns. Details on selection cases, division model development validation data raw frequently ambiguous missing. as having areas but requires more rigorous evaluation its performance.

Language: Английский

---