EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Automated lesion detection in endoscopic imagery for small animal models – a pilot study

Thomas Eixelberger iD, Ralf Hackner, Qi Fang, Bisan Abdalfatah Zohud, Michael Stürzl iD, Elisabeth Naschberger iD, Thomas Wittenberg iD

Biomedical Engineering / Biomedizinische Technik · 2025

DOI: 10.1515/bmt-2025-0179 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Biomedical Engineering / Biomedizinische Technik, 2025. Verfasst von Thomas Eixelberger und 6 weiteren Autor:innen. Beteiligte Einrichtung: Chair of Visual Computing, Friedrich-Alexander-University Erlangen-Nürnberg, Cauerstr. 11, 91058 Erlangen, Germany. Hauptthemen laut MEDLINE: Colonoscopy, Colorectal Neoplasms, Colonic Neoplasms, Image Processing, Computer-Assisted. Die Arbeit nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Deutsche Forschungsgemeinschaft.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Objectives

Small animal models, particularly mice, are crucial for studying gastrointestinal diseases like colorectal cancer. Tumor assessment via colonoscopy generates large video datasets, necessitating automated analysis due to limited resources and time-consuming manual review.

Methods

We employed a YOLOv7-based deep learning model pre-trained on human polyp images to detect tumors in mouse colonoscopy videos. Detection was enhanced using a stool detector and a color-based filter. Lesions were classified from ‘0’ (no tumor) to ‘5’ (tumor >50 % of colon diameter) using a custom ratio-based method. The system was evaluated on 150 videos from 28 mice over 6 weeks, with 125 videos containing tumors.

Results

Initial detection yielded a Precision of 0.576, Recall of 0.916, and Accuracy of 0.593. Adding the stool detector improved results to 0.932, 0.946, and 0.897, respectively. Compared to expert annotations, classification reached 0.759 Precision, 0.774 Recall, and 0.774 Accuracy over all five classes.

Conclusions

The proposed approach reliably detects and classifies colon tumors in mice, offering real-time support for preclinical endoscopic studies. Further evaluation will provide more insights into its performance.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

ColonoscopyColorectal NeoplasmsColonic NeoplasmsImage Processing, Computer-AssistedAnimalsMicePilot ProjectsHumansDeep LearningColonic Polyps

Stichwörter der Autor:innen

YOLOv7-tinyautomated lesion detectioncancer researchcolorectal cancermouse modelstumor grading

Autor:innen und Institutionen

  • Chair of Visual Computing, Friedrich-Alexander-University Erlangen-Nürnberg, Cauerstr. 11, 91058 Erlangen, Germany.
  • Digital Health and Analytics, Fraunhofer Institute for Integrated Circuits IIS, Am Wolfsmantel 33, 91058 Erlangen, Germany.
  • Department of Surgery, Universitätsklinikum Erlangen, Schwabachanlage 12, 91054 Erlangen, Germany.

ORCID-Profile: Thomas Eixelberger, Michael Stürzl, Elisabeth Naschberger, Thomas Wittenberg

Förderung

  • Deutsche Forschungsgemeinschaft

Zitieren

APA

Eixelberger, T., Hackner, R., Fang, Q., Zohud, B. A., Stürzl, M., Naschberger, E., & Wittenberg, T. (2025). Automated lesion detection in endoscopic imagery for small animal models – a pilot study. Biomedical Engineering / Biomedizinische Technik. https://doi.org/10.1515/bmt-2025-0179

Vancouver

Eixelberger T, Hackner R, Fang Q, Zohud BA, Stürzl M, Naschberger E, et al. Automated lesion detection in endoscopic imagery for small animal models – a pilot study. Biomedical Engineering / Biomedizinische Technik. 2025. doi:10.1515/bmt-2025-0179

BibTeX
@article{Eixelberger2025,
  author = {Eixelberger, Thomas and Hackner, Ralf and Fang, Qi and Zohud, Bisan Abdalfatah and Stürzl, Michael and Naschberger, Elisabeth and Wittenberg, Thomas},
  title = {Automated lesion detection in endoscopic imagery for small animal models – a pilot study},
  journal = {Biomedical Engineering / Biomedizinische Technik},
  year = {2025},
  doi = {10.1515/bmt-2025-0179}
}
Kostenlos starten →