Performance of an Artificial Intelligence Program for Assessing the Depth of Early Gastric Cancer Using Endoscopic Images
Leistung eines KI-Programms zur Beurteilung der Tiefe des frühen Magenkarzinoms anhand endoskopischer Bilder
Endoscopy International Open · 17.09.2026 · 14 (CP)
DOI: 10.1055/a-2957-0692 ↗Auf einen Blick
Ein KI-Programm wurde mit endoskopischen Bildern von 204 Fällen getestet. Die Sensitivität für mukosales Karzinom lag bei 91,2 %; die Untergrenze des 95 %-KI lag über 75 %.
Fragestellung
Wie gut diagnostiziert ein KI-Programm anhand von Endoskopiebildern die Tiefe eines frühen Magenkarzinoms?
Methode
Testdatensatz mit 204 Fällen aus 2018 bis 2021; mindestens fünf WLI-Bilder pro Fall; primärer Endpunkt Sensitivität für mukosales Karzinom.
Ergebnisse
Sensitivität 91,2 %; Untergrenze des 95 %-KI 85,2 %; Spezifität 66,2 %; Genauigkeit 82,8 %. Die Untergrenze lag über 75 % und erfüllte damit das Kriterium für Nützlichkeit.
Schlussfolgerung der Autor:innen
Das entwickelte KI-Programm zur Diagnose der Tiefe des frühen Magenkarzinoms ist nützlich und dürfte klinisch anwendbar sein.
Kernaussagen
- Sensitivität für mukosales Karzinom: 91,2 %
- Untergrenze des 95 %-KI: 85,2 %
- Spezifität 66,2 %; Genauigkeit 82,8 %
Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.
Kurz gefasst
Fachartikel in der Zeitschrift Endoscopy International Open, erschienen am 17.09.2026. Verfasst von Shoichiro Hirata und 15 weiteren Autor:innen. Beteiligte Einrichtung: Department of Gastroenterology and HepatologyOkayama University Graduate School of Medicine, Dentistry and Pharmaceutical SciencesOkayamaJapan. Die Arbeit nennt 17 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Background and Study Aims The depth of early gastric cancer is an important factor in determining treatment plans, such as endoscopic resection or surgery. We developed an artificial intelligence (AI) program using deep neural networks to diagnose mucosal and submucosal cancers via endoscopic images. Performance evaluation tests were conducted for its clinical applications. Patients and Methods A test dataset comprising 204 early gastric cancer cases treated with endoscopic resection or surgery at our hospital between 2018 and 2021 was established. A minimum of five white-light imaging (WLI) images per case were used (median, 8; range, 5–22). The primary endpoint was the sensitivity for diagnosing mucosal cancer. The AI program was defined as useful when the lower limit of the 95% confidence interval (CI) exceeded 75%, based on previous reports on the depth diagnosis of early gastric cancer. Secondary endpoints were specificity and accuracy.
Results
In our study, the sensitivity of this program for the diagnosis of mucosal cancer was 91.2%, and the lower limit of the 95% CI of sensitivity was 85.2%, which was higher than 75%, indicating the usefulness of this program. The specificity and accuracy were 66.2% and 82.8%, respectively.
Conclusions
The AI program developed to diagnose the depth of early gastric cancer is useful and expected to be clinically applicable.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Gastroenterology and HepatologyOkayama University Graduate School of Medicine, Dentistry and Pharmaceutical SciencesOkayamaJapan.
- Department of Practical Gastrointestinal Endoscopy, Faculty of MedicineOkayama University Graduate School of Medicine, Dentistry and Pharmaceutical SciencesOkayamaJapan.
- Department of Regenerative Medicine, Center for Innovative Clinical MedicineOkayama University HospitalOkayamaJapan.
- Department of Gastroenterological SurgeryOkayama University Graduate School of Medicine Dentistry and Pharmaceutical SciencesOkayamaOkayamaJapan.
- Department of PathologyOkayama University HospitalOkayamaJapan.
ORCID-Profile: Masaya Iwamuro
Zitieren
APA
Hirata, S., Hamada, K., Kawahara, Y., Aya, Y., Nishio, A., Yoshikawa, T., Maki, Y., Satomi, T., Matsueda, K., Sakaguchi, C., Kono, Y., Uchida, D., Iwamuro, M., Kuroda, S., Tanaka, T., & Otsuka, M. (2026). Performance of an Artificial Intelligence Program for Assessing the Depth of Early Gastric Cancer Using Endoscopic Images. Endoscopy International Open, 14(CP). https://doi.org/10.1055/a-2957-0692
Vancouver
Hirata S, Hamada K, Kawahara Y, Aya Y, Nishio A, Yoshikawa T, et al. Performance of an Artificial Intelligence Program for Assessing the Depth of Early Gastric Cancer Using Endoscopic Images. Endoscopy International Open. 2026;14(CP). doi:10.1055/a-2957-0692
BibTeX
@article{Hirata2026,
author = {Hirata, Shoichiro and Hamada, Kenta and Kawahara, Yoshiro and Aya, Yusuke and Nishio, Ayano and Yoshikawa, Tomoki and Maki, Yoshie and Satomi, Takuya and Matsueda, Katsunori and Sakaguchi, Chihiro and Kono, Yoshiyasu and Uchida, Daisuke and Iwamuro, Masaya and Kuroda, Shinji and Tanaka, Takehiro and Otsuka, Motoyuki},
title = {Performance of an Artificial Intelligence Program for Assessing the Depth of Early Gastric Cancer Using Endoscopic Images},
journal = {Endoscopy International Open},
year = {2026},
volume = {14},
number = {CP},
doi = {10.1055/a-2957-0692}
}
EUVIMED