Abstract
Current gastric cancer (GCa) risk systems are prone to errors since they evaluate a visual estimation of intestinal metaplasia percentages in histopathology images of gastric mucosa to assign a risk. This study presents an automated method to detect and quantify intestinal metaplasia using deep convolutional neural networks as well as a comparative analysis with visual estimations of three pathologists. Gastric samples were collected from two different cohorts: 149 asymptomatic volunteers from a region with a high prevalence of GCa in Colombia and 56 patients from a tertiary hospital. Deep learning models were trained to classify intestinal metaplasia, and predictions were used to estimate a percentage of intestinal metaplasia and to assign an adapted OLGIM stage. Atrophy was not assessed because of the limited reproducibility among pathologists. Results were compared with independent blinded metaplastic assessments performed by three graduated pathologists. The best-performing deep learning architecture classified intestinal metaplasia with F1-Score of \(0.80 \pm 0.01\) and AUC of \(0.91 \pm 0.01\). Among pathologists, inter-observer agreement by a Fleiss’s Kappa score ranged from 0.20 to 0.48. In comparison, agreement between the pathologists and the best-performing model ranged from 0.12 to 0.35. Deep learning models show potential to reliably detect and quantify the percentage of intestinal metaplasia, achieving high classification performance. In practice, visual estimation is still the only available method, yet it is marked by considerable inter-observer variability. Deep learning models provide consistent estimates that could help reduce this subjectivity in risk stratification.
Data Availability
The collection of internal gastric samples used in this study were obtained from the primary research project entitled Investigación de la prevalencia de lesiones precursoras de malignidad gástrica y efecto de la erradicación de Helicobacter pylori colon prevención primaria de cáncer gástrico en el Departamento de Nariño. This project was approved under Agreement No. 057 of 2017 by the Collegiate Body for Administration and Decision (OCAD Pacífico), and was funded by the Science, Technology, and Innovation Fund of the General System of Royalties/Government of Nariño. Internal and external data are publicly available at Harvard Dataverse59.
References
Bray, F. et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J. Clin. 74(3), 229–263 (2022).
Ajani, J. et al. Gastric adenocarcinoma. Nat. Rev. Dis. Prim. 3, 1–19 (2017).
Tan, P. & Yeoh, K. Genetics and molecular pathogenesis of gastric adenocarcinoma. Gastroenterology 149, 1153–1162 (2015).
Pucułek, M. et al. Helicobacter pylori associated factors in the development of gastric cancer with special reference to the early-onset subtype. Oncotarget 9(57), 31146–31162 (2018).
Cai, Q. et al. Development and validation of a prediction rule for estimating gastric cancer risk in the Chinese high-risk population: a nationwide multicentre study. Gut 68(9), 1576–1587 (2019).
Misiewicz, J. The Sydney System: a new classification of gastritis. Introduction. J. Gastroenterol. Hepatol. 6(3), 207–208 (1991).
Dixon, M. et al. Classification and grading of gastritis. The updated Sydney System. International Workshop on the Histopathology of Gastritis, Houston 1994. Am. J. Surg. Pathol. 20, 1161–1181 (1996).
Rugge, M. et al. Gastritis staging in clinical practice: the olga staging system. Gut 56(5), 631–636 (2007).
Crafa, P. et al. From Sidney to OLGA: an overview of atrophic gastritis. Acta Biomed. 89(8–S), 93–99 (2018).
Correa, P. Gastric cancer: overview. Gastroenterol. Clin. N. Am. 42(2), 211–217 (2013).
Koulis, A. et al. Premalignant lesions and gastric cancer: Current understanding. World J. Gastrointest. Oncol. 11(9), 665–678 (2019).
Tjandra, D. et al. Gastric intestinal metaplasia: Challenges and the opportunity for precision prevention. Cancers 15, 3913 (2023).
Marcos, P. et al. Endoscopic grading of gastric intestinal metaplasia on risk assessment for early gastric neoplasia: can we replace histology assessment also in the west?. Gut 69(10), 1762–1768 (2020).
Rugge, M. & Genta, R. Staging and grading of chronic gastritis. Hum. Pathol. 36(3), 228–233 (2005).
Capelle, L. et al. The staging of gastritis with the OLGA system by using intestinal metaplasia as an accurate alternative for atrophic gastritis. Gastrointest. Endosc. 71(7), 1150–1158 (2010).
Fang, S. et al. Diagnosing and grading gastric atrophy and intestinal metaplasia using semi-supervised deep learning on pathological images: development and validation study. Gastr. Cancer 27, 343–354 (2024).
Zhao, Q. et al. Deep learning model can improve the diagnosis rate of endoscopic chronic atrophic gastritis: a prospective cohort study. BMC Gastroenterol. 22(1), 133 (2022).
Yue, H. et al. The significance of OLGA and OLGIM staging systems in the risk assessment of gastric cancer: a systematic review and meta-analysis. Gastr. Cancer 21, 579–587 (2018).
Isajevs, S. et al. Gastritis staging: interobserver agreement by applying OLGA and OLGIM systems. Virchows Arch. 464, 403–407 (2014).
Molaei, M. et al. Gastric atrophy: use of OLGA staging system in practice. Gastroenterol. Hepatol. Bed. Bench. 9(1), 25–29 (2016).
Rugge, M. et al. Gastritis OLGA-staging and gastric cancer risk: a twelve-year clinicopathological followup study. Aliment. Pharmacol. Ther. 31, 1104–1111 (2010).
Mansour-Ghanaei, F. et al. OLGA- and OLGIM-based staging in the patients with gastritis and endoscopy indications. Turk. J. Gastroenterol. 33(2), 95–102 (2022).
Chen, C. et al. An annotation-free whole-slide training approach to pathological classification of lung cancer types using deep learning. Nat. Commun. 12, 1193 (2021).
Coudray, N. et al. Classification and mutation prediction from non–small cell lung cancer histopathology images using deep learning. Nat. Med. 24, 1559–1567 (2018).
Kanavati, F. et al. Weakly-supervised learning for lung carcinoma classification using deep learning. Sci. Rep. 10, 9297 (2020).
Cruz-Roa, A. et al. High-throughput adaptive sampling for whole-slide histopathology image analysis (HASHI) via convolutional neural networks: Application to invasive breast cancer detection. PLoS One 13(5), e0196828 (2018).
Liu, M. et al. A deep learning method for breast cancer classification in the pathology images. IEEE J. Biomed. Health Inform. 26(10), 5025–5032 (2022).
Wetstein, S. et al. Deep learning-based breast cancer grading and survival analysis on whole-slide histopathology images. Sci. Rep. 12, 15102 (2022).
Nir, G. et al. Automatic grading of prostate cancer in digitized histopathology images: Learning from multiple experts. Med. Image Anal. 50, 167–180 (2018).
Li, Y. et al. Automated gleason grading and gleason pattern region segmentation based on deep learning for pathological images of prostate cancer. IEEE Access. 8, 117714–117725 (2020).
Tolkach, Y. et al. High-accuracy prostate cancer pathology using deep learning. Nat. Mach. Intell. 2, 411–418 (2020).
Wang, X. et al. Predicting gastric cancer outcome from resected lymph node histopathology images using deep learning. Nat. Commun. 12, 1637 (2021).
Huang, B. et al. Accurate diagnosis and prognosis prediction of gastric cancer using deep learning on digital pathological images: A retrospective multicentre study. EBioMedicine. 73, 103631 (2021).
Sharma, H. et al. Deep convolutional neural networks for automatic classification of gastric carcinoma using whole slide images in digital histopathology. Comput. Med. Imaging Graph. 61, 2–13 (2017).
Veldhuizen, G. P. et al. Deep learning-based subtyping of gastric cancer histology predicts clinical outcome: a multi-institutional retrospective study. Gastr. Cancer 26(5), 708–720 (2023).
White, J. et al. Identifying the pre-malignant stomach: from guidelines to practice. Transl. Gastroenterol. Hepatol. 7, 8 (2022).
Pimentel-Nunes, P. et al. Management of epithelial precancerous conditions and lesions in the stomach (MAPS II): European Society of Gastrointestinal Endoscopy (ESGE), European Helicobacter and Microbiota Study Group (EHMSG), European Society of Pathology (ESP), and Sociedade Portuguesa de Endoscopia Digestiva (SPED) guideline update 2019. Endoscopy 51, 365–388 (2019).
Hwang, Y. et al. Reversibility of atrophic gastritis and intestinal metaplasia after helicobacter pylori eradication—A prospective study for up to 10 years. Aliment. Pharmacol. Ther. 47, 380–390 (2018).
Aumpan, N. et al. Predictors for regression and progression of intestinal metaplasia (IM): A large population-based study from low prevalence area of gastric cancer (IM-predictor trial). PLoS One 16(8), e0255601 (2021).
Bedoya, A. et al. Proyecto Urkunina 5000. Investigación de la prevalencia de lesiones precursoras y del efecto de la erradicación de Helicobacter pylori como prevención primaria del cáncer gástrico en el departamento de Nariño. Rev. Colomb. Cir. 33(4), 345–352 (2019).
Sanderson, R. et al. Web Annotation Data Model: W3C Recommendation 23 February 2017. In Proceedings of the 5th Annual ACM Web Science Conference, 366–375 (2013).
Sirinukunwattana, K. et al. Gland segmentation in colon histology images: The glas challenge contest. Med. Image Anal. 35, 489–502 (2017).
Jass, J. Classification of colorectal cancer based on correlation of clinical, morphological and molecular features. Histopathology 50(1), 113–130 (2007).
Shah, S. C. et al. Histologic subtyping of gastric intestinal metaplasia: overview and considerations for clinical practice. Gastroenterology 158(3), 745–750 (2020).
Deng, J. et al. Imagenet: A large-scale hierarchical image database. In 2009 IEEE Conference on Computer Vision and Pattern Recognition, 248–255 (2009).
He, K. et al. Deep residual learning for image recognition. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 770–778 (2016).
Huang, G. et al. Densely connected convolutional networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 4700–4708 (2017).
Liu, Z. et al. A convnet for the 2020s. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 11976–11986 (2021).
Caviedes, M. et al. An automatic classification of metaplasia in gastric histopathology images. In 2023 19th International Symposium on Medical Information Processing and Analysis (SIPAIM), 1–4 (2023).
Riasatian, A. et al. Fine-tuning and training of densenet for histopathology image representation using tcga diagnostic slides. Med. Image Anal. 70, 102032 (2021).
Bravo, D. et al. Automatic classification of esophagogastroduodenoscopy sub-anatomical regions. In 2023 IEEE 20th International Symposium on Biomedical Imaging (ISBI), 1–5 (2023).
Chen, R. J. et al. Towards a general-purpose foundation model for computational pathology. Nat. Med. 30, 850–862 (2024).
Tomczak, K. et al. The cancer genome atlas (TCGA): an immeasurable source of knowledge. Contemp Oncol (Pozn). 19(1A), A68-77 (2015).
Lerch, J. M. et al. Subtyping intestinal metaplasia in patients with chronic atrophic gastritis: an interobserver variability study. Pathology 54(3), 262–268 (2022).
Salazar, B. et al. The OLGA-OLGIM staging and the interobserver agreement for gastritis and preneoplastic lesion screening: a cross-sectional study. Virchows Arch. 480, 759–769 (2022).
Selvaraju, R. et al. Grad-CAM: Visual explanations from deep networks via gradient-based localization. Proc. IEEE Int. Conf. Comput. Vis. ICCV 2017, 618–626 (2017).
Rugge, M. et al. Operative link for gastritis assessment vs operative link on intestinal metaplasia assessment. World J. Gastroenterol. 17(41), 4596–4601 (2011).
Janowczyk, A. & Madabhushi, A. Deep learning for digital pathology image analysis: A comprehensive tutorial with selected use cases. J. Pathol. Inform. 7(1), 29 (2016).
Cano, F. et al. Supplemented data for: Towards deep-learning based detection and quantification of intestinal metaplasia on digitized gastric biopsies: a multi-expert comparative study. Harvard Dataverse. https://doi.org/10.7910/DVN/NVUCW8 (2025).
Acknowledgements
Special thanks to the CIEDYN foundation and the project BPIN 20150000100064 Urkunina5000, from which whole slides of gastric tissue were recovered from asymptomatic volunteers and subsequently digitized to be used in the development of this work.
Funding
This work was partially supported by the project with code 110192092345 Program for the Early Detection of Premalignant Lesions and Gastric Cancer in urban, rural and dispersed areas in the Department of Nariño of call No. 920 of 2022 of MinCiencias. This work was partially supported by project 52895, titled Proposal for the strategic plan for the establishment of the Center of Excellence (Inter-Sites) in Medicine and Artificial Intelligence (SemAI), from the National Call for Proposals Bank for the Consolidation of Centers of Excellence 2020-2021 at Universidad Nacional de Colombia. This work was partially supported by project BPIN 2019000100060 Implementation of a Network for Research, Technological Development and Innovation in Digital Pathology (RedPat) supported by Industry 4.0 technologies from FCTeI of SGR resources, which was approved by OCAD of FCTeI and MinCiencias.
Author information
Authors and Affiliations
Contributions
Conceptualization: F.C., A.C.R., F.A.G., S.E.V. and E.R.; Data curation: F.C., M.C., A.S., J.V. and J.Q.; Methodology: F.C., M.C., A.C.R., F.A.G. and E.R.; Formal analysis: F.C., A.C.R., F.A.G., S.E.V. and E.R.; Writing-original draft preparation: F.C. and E.R.; Writing-review and editing: F.C., A.C.R., F.A.G., S.E.V. and E.R.; Funding acquisition: A.C.R. and E.R.; Resources: A.B.U., M.C.B., Y.Y.C., A.C.R., F.A.G. and E.R.; Supervision: F.A.G., S.E.V. and E.R.; Project administration: E.R. All authors have read and agreed to the published version of the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Ethical considerations
This research study was conducted retrospectively using data provided by the CIEDYN foundation, a partner of the Urkunina5000 project, which contains information on ethical considerations in compliance with the Declaration of Helsinki. All patients signed informed consent forms. Additional ethics considerations was not required.
Informed consent
All asymptomatic volunteers selected for this study signed informed consent and all guarantees of anonymization were applied to their data.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Cano, F., Caviedes, M., Siabatto, A. et al. Towards deep-learning based detection and quantification of intestinal metaplasia on digitized gastric biopsies: a multi-expert comparative study. Sci Rep (2026). https://doi.org/10.1038/s41598-025-32737-w
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-025-32737-w
