The use of artificial intelligence methods in pathology

Czech version

Authors: Vojtěch Damian; Uladzislau Palichenka; Miroslav Koblížek; Petr Škapa; Josef Zámečník
Authors‘ workplace: Ústav patologie a molekulární medicíny, Fakultní nemocnice v Motole, 2. lékařská fakulta Univerzity Karlovy, Praha
Published in: Čes-slov Pediat 2025; 80 (5): 226-230.
Category:
doi: https://doi.org/10.55095/cspediatrie2025/043

Overview

Damian V, Palichenka U, Koblížek M, Škapa P, Zámečník J. The use of artificial intelligence methods in pathology

Modern pathology is currently undergoing a fundamental transformation. It is evolving from a purely morphological discipline to an integrated field that combines histopathology with molecular biology and bioinformatics. Rapid advances in computational technology and artificial intelligence (AI) have enabled the processing of not only data generated by molecular pathology analyzers, but also of microscopic images, produced by high-capacity slide scanners.

These digitized microscopic slides, also known as whole slide images (WSI), can then be algorithmically processed using a wide range of machine learning tools, including deep learning, a subset of machine learning based on artificial neural networks. These deep learning-based approaches have opened new possibilities in diagnostics. They have demonstrated the ability to, among others, detect tumors and lymph node metastases, estimate tumor grade, and even predict molecular alterations.

As in other disciplines of medicine, digitization in pathology is not without its challenges. The preparation of histological slides remains largely a manual process making it essential to maintain high-quality standards, as scanners and digital pathology tools rely heavily on clear, artefact-free slides to function properly. In addition to these pathology-specific limitations, broader economic, legal, and technical challenges must also be addressed, including the storage and analysis of massive volumes of data, which may reach hundreds or even thousands of terabytes annually.

Digital and computational pathology are rapidly advancing fields with profound implications for the future of cancer diagnostics and personalized medicine.

Keywords:

artificial intelligence – deep learning – Digital Pathology – computational pathology

Sources

1. OpenAI, Achiam J, Adler S, et al. GPT-4 Technical Report. 2024. doi: 10.48550/arXiv.2303.08774

2. Chen RJ, Ding T, Lu MY, et al. Towards a general-purpose foundation model for computational pathology. Nat Med 2024; 30(3): 850–862. doi: 10.1038/s41591-024-02857-3

3. Unger M, Kather JN. Deep learning in cancer genomics and histopathology. Genome Med 2024; 16(1): 44. doi: 10.1186/s13073-024-01315-6

4. Matthews GA, McGenity C, Bansal D, Treanor D. Public evidence on AI products for digital pathology. Npj Digit Med 2024; 7(1): 300. doi: 10.1038/s41746-024-01294-3

5. Ertosun MG, Rubin DL. Automated grading of gliomas using deep learning in digital pathology images: a modular approach with ensemble of convolutional neural networks. AMIA Annu Symp Proc AMIA Symp 2015 : 1899–1908.

6. Kather JN, Pearson AT, Halama N, et al. Deep learning can predict microsatellite instability directly from histology in gastrointestinal cancer. Nat Med 2019; 25(7): 1054–1056. doi: 10.1038/s41591-019-0462-y

7. Fremond S, Andani S, Wolf JB, et al. Interpretable deep learning model to predict the molecular classification of endometrial cancer from haematoxylin and eosin-stained whole-slide images: a combined analysis of the PORTEC randomised trials and clinical cohorts. Lancet Digit Health 2023; 5(2): e71–e82. doi: 10.1016/S2589-7500(22)00210-2

8. Jaber MI, Song B, Taylor C, et al. A deep learning image-based intrinsic molecular subtype classifier of breast tumors reveals tumor heterogeneity

that may affect survival. Breast Cancer Res 2020; 22(1): 12. doi: 10.1186/s13058-020-1248-3

9. Fu Y, Jung AW, Torne RV, et al. Pan-cancer computational histopathology reveals mutations, tumor composition and prognosis. Nat Cancer 2020; 1(8): 800–810. doi: 10.1038/s43018-020-0085-8

10. Vorontsov E, Bozkurt A, Casson A, et al. A foundation model for clinical-grade computational pathology and rare cancers detection. Nat Med 2024; 30(10): 2924–2935. doi: 10.1038/s41591-024-03141-0

11. Oquab M, Darcet T, Moutakanni T, et al. DINOv2: Learning Robust Visual Features without Supervision. 2024. doi: 10.48550/arXiv.2304.07193

12. Touvron H, Martin L, Stone K, et al. Llama 2: open foundation and fine-tuned chat models. 2023. doi: 10.48550/arXiv.2307.09288

13. Lu MY, Chen B, Williamson DFK, et al. A multimodal generative AI copilot for human pathology. Nature 2024; 634(8033): 466–473. doi: 10.1038/s41586-024-07618-3

14. PathChat Receives FDA Breakthrough Device Designation | Modella AI. Dostupné na: https: //www.modella.ai/pathchat-fda-breakthrough-designation

15. Zeng Y, Wei Z, Yu W, et al. Spatial transcriptomics prediction from histology jointly through Transformer and graph neural networks. Brief Bioinform 2022; 23(5): bbac297. doi: 10.1093/bib/bbac297

16. McLendon R, Friedman A, Bigner D, et al. Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature 2008; 455(7216): 1061–1068. doi: 10.1038/nature07385