EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Challenges in Implementing Endoscopic Artificial Intelligence: The Impact of Real‐World Imaging Conditions on Barrett's Neoplasia Detection

M R Jong, T J M Jaspers, C H J Kusters, J B Jukema, R A H van Eijck van Heslinga, K N Fockens, T G W Boers, L S Visser, J A van der Putten, F van der Sommen iD, P H de With, A J de Groof iD und 2 weitere

United European Gastroenterology Journal · 2025

DOI: 10.1002/ueg2.12760 ↗

Kurz gefasst

Fachartikel in der Zeitschrift United European Gastroenterology Journal, 2025. Verfasst von M R Jong und 13 weiteren Autor:innen. Beteiligte Einrichtung: Department of Gastroenterology and Hepatology, Amsterdam Gastroenterology, Endocrinology and Metabolism, Amsterdam UMC, University of Amsterdam, Amst…. Hauptthemen laut MEDLINE: Barrett Esophagus, Esophageal Neoplasms, Esophagoscopy, Deep Learning. Die Arbeit wurde bisher 9-mal zitiert (Crossref) und nennt 23 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Nederlandse Organisatie voor Wetenschappelijk Onderzoek.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Endoscopic deep learning systems are often developed using high‐quality imagery obtained from expert centers. Therefore, they may underperform in community hospitals where image quality is more heterogeneous.

Objective

This study aimed to quantify the performance degradation of a computer aided detection system for Barrett's neoplasia, trained on expert images, when exposed to more heterogeneous imaging conditions representative of daily clinical practice. Further, we evaluated strategies to mitigate this performance loss.

Methods

We developed a computer aided detection system using 1011 high‐quality, expert‐acquired images from 373 Barrett's patients. We assessed its performance on high, moderate and low image quality test sets, each containing images from an independent group of 117 Barrett's patients. These test sets reflected the varied image quality of routine patient care and contained artefacts such as insufficient mucosal cleaning and inadequate esophageal expansion. We then applied three methods to improve the algorithm's robustness to data heterogeneity: inclusion of more diverse training data, domain‐specific pretraining and architectural optimization.

Results

The computer aided detection system, when trained exclusively on high‐quality data, achieved area under the curve (AUC), sensitivity and specificity scores of 83%, 85% and 67% on the high quality test set. AUC and sensitivity were significantly lower with 80% ( p < 0.001) and 62% ( p = 0.002) on the moderate‐quality and 71% ( p > 0.001) and 47% ( p = 0.002) on the low‐quality test set. Incorporating robustness‐enhancing strategies significantly improved the AUC, sensitivity and specificity to 92% ( p = 0.004), 88% ( p = 0.84) and 81% ( p = 0.003) on the high‐quality test set, 93% ( p = 0.006), 86% ( p = 0.01) and 83% ( p = 0.09) on the moderate‐quality test set and 84% ( p = 0.001), 78% ( p = 0.002) and 77% ( p = 0.23) on the low‐quality test set.

Conclusion

Endoscopic deep learning systems trained solely on high‐quality images may not perform well when exposed to heterogeneous imagery, as found in routine practice. Robustness‐enhancing training strategies can increase the likelihood of successful clinical implementation.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Barrett EsophagusEsophageal NeoplasmsEsophagoscopyDeep LearningArtificial IntelligenceImage Interpretation, Computer-AssistedDiagnosis, Computer-AssistedHumansFemaleMaleSensitivity and SpecificityAlgorithmsMiddle AgedAgedEsophagus

Stichwörter der Autor:innen

Barrett's esophagusartificial intelligencecomputer aided detectiondeep learning systemsendoscopyesophageal adenocarcinoma

Autor:innen und Institutionen

  • Department of Gastroenterology and Hepatology, Amsterdam Gastroenterology, Endocrinology and Metabolism, Amsterdam UMC, University of Amsterdam, Amsterdam, The Netherlands.
  • Department of Electrical Engineering, Eindhoven University of Technology, Eindhoven, The Netherlands.

ORCID-Profile: F van der Sommen, A J de Groof, J J Bergman

Förderung

  • Nederlandse Organisatie voor Wetenschappelijk Onderzoek

Zitieren

APA

Jong, M. R., Jaspers, T. J. M., Kusters, C. H. J., Jukema, J. B., van Eijck van Heslinga, R. A. H., Fockens, K. N., Boers, T. G. W., Visser, L. S., van der Putten, J. A., van der Sommen, F., de With, P. H., de Groof, A. J., Bergman, J. J., & BONS‐AI consortium (2025). Challenges in Implementing Endoscopic Artificial Intelligence: The Impact of Real‐World Imaging Conditions on Barrett's Neoplasia Detection. United European Gastroenterology Journal. https://doi.org/10.1002/ueg2.12760

Vancouver

Jong MR, Jaspers TJM, Kusters CHJ, Jukema JB, van Eijck van Heslinga RAH, Fockens KN, et al. Challenges in Implementing Endoscopic Artificial Intelligence: The Impact of Real‐World Imaging Conditions on Barrett's Neoplasia Detection. United European Gastroenterology Journal. 2025. doi:10.1002/ueg2.12760

BibTeX
@article{Jong2025,
  author = {Jong, M R and Jaspers, T J M and Kusters, C H J and Jukema, J B and van Eijck van Heslinga, R A H and Fockens, K N and Boers, T G W and Visser, L S and van der Putten, J A and van der Sommen, F and de With, P H and de Groof, A J and Bergman, J J and BONS‐AI consortium, BONS‐AI consortium},
  title = {Challenges in Implementing Endoscopic Artificial Intelligence: The Impact of Real‐World Imaging Conditions on Barrett's Neoplasia Detection},
  journal = {United European Gastroenterology Journal},
  year = {2025},
  doi = {10.1002/ueg2.12760}
}
Kostenlos starten →