EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Artificial intelligence for TNM staging in NSCLC: a critical appraisal of segmentation utility in [1⁸F]FDG PET/CT

Maurice M Heimer iD, Jakob Dexl, Johanna Ta, Ricarda Ebner, Felix L Herr, Leon Orasanin, Katharina Jeblick, Lisa C Adams, Lalith K Shiyam Sundar, Amanda Tufman, Rudolf A Werner, Gabriel Sheikh und 4 weitere

European Journal of Nuclear Medicine and Molecular Imaging · 2026

DOI: 10.1007/s00259-025-07677-2 ↗

Kurz gefasst

Fachartikel in der Zeitschrift European Journal of Nuclear Medicine and Molecular Imaging, 2026. Verfasst von Maurice M Heimer und 15 weiteren Autor:innen. Beteiligte Einrichtung: Department of Radiology, LMU University Hospital, LMU Munich, Munich, Germany. maurice.heimer@med.uni-muenchen.de. Hauptthemen laut MEDLINE: Positron Emission Tomography Computed Tomography, Fluorodeoxyglucose F18, Carcinoma, Non-Small-Cell Lung, Lung Neoplasms. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 29 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Klinikum der Universität München.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Purpose

This study aims to investigate whether a diagnostic AI model can effectively support lesion detection and staging in non-small cell lung cancer (NSCLC) [ 1 ⁸F]FDG PET/CT studies, focusing on the distinction between technical segmentation accuracy and clinically meaningful performance.

Methods

In this retrospective single-centre study, [ 1 ⁸F]FDG PET/CT scans from 306 treatment-naïve NSCLC patients were reviewed with reference to multidisciplinary team decisions. Tumour lesions were manually segmented for reference and compared with predictions from the top-performing algorithm of the autoPET III challenge. Quantitative segmentation metrics were calculated, and lesion-level errors were assessed for impact on patient-level TNM and UICC staging.

Results

The algorithm achieved a mean Dice Similarity Coefficient (DSC) of 0.64. Lesion-level sensitivity was 95.8% across all patients, with a precision of 87.5%. False positive M-category lesions ( n = 196) occurred as most frequent error. Of all false positives, 35.7% were benign and 34.7% non-oncologic pathologies. UICC staging matched ground truth in 207/306 patients, with most discordances due to upstaging (88/306).

Conclusion

Clinically driven metrics and cause-based error analysis offer valuable insight into AI segmentation performance. The evaluated model showed excellent lesion sensitivity but a tendency towards systematic overprediction across TNM categories. On a lesion level M-stage false positives and undersegmentation in the hilar region emerged as the main driver of clinically relevant upstaging. Despite promising lesion detection sensitivity, only 67.7% UICC-stagings were accurate using AI masks, indicating that diagnostic AI may support, though not yet replace, manual lesion evaluation in NSCLC [ 1 ⁸F]FDG PET/CT.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Positron Emission Tomography Computed TomographyFluorodeoxyglucose F18Carcinoma, Non-Small-Cell LungLung NeoplasmsArtificial IntelligenceImage Processing, Computer-AssistedHumansNeoplasm StagingFemaleMaleMiddle AgedAgedRetrospective StudiesAged, 80 and overAdultRadiopharmaceuticals

Stichwörter der Autor:innen

Artificial intelligenceExternal validationNon-small cell lung cancerTask-specific evaluation[¹⁸F] FDG PET/CT

Substanzen

Fluorodeoxyglucose F18Radiopharmaceuticals

Autor:innen und Institutionen

  • Department of Radiology, LMU University Hospital, LMU Munich, Munich, Germany. maurice.heimer@med.uni-muenchen.de.
  • Department of Radiology, LMU University Hospital, LMU Munich, Munich, Germany.
  • Munich Center for Machine Learning (MCML), Munich, Germany.
  • Comprehensive Pneumology Center (CPC-M), Member of the German Center for Lung Research (DZL), Munich, Germany.
  • Department of Radiology, TUM University Hospital, TU Munich, Munich, Germany.
  • Department of Medicine V, LMU University Hospital, LMU Munich, Munich, Germany, and Bavarian Center for Cancer Research (BZKF), Munich, Germany.

ORCID-Profile: Maurice M Heimer

Förderung

  • Klinikum der Universität München

Zitieren

APA

Heimer, M. M., Dexl, J., Ta, J., Ebner, R., Herr, F. L., Orasanin, L., Jeblick, K., Adams, L. C., Sundar, L. K. S., Tufman, A., Werner, R. A., Sheikh, G., Ricke, J., Ingrisch, M., Fabritius, M. P., & Cyran, C. C. (2026). Artificial intelligence for TNM staging in NSCLC: a critical appraisal of segmentation utility in [1⁸F]FDG PET/CT. European Journal of Nuclear Medicine and Molecular Imaging. https://doi.org/10.1007/s00259-025-07677-2

Vancouver

Heimer MM, Dexl J, Ta J, Ebner R, Herr FL, Orasanin L, et al. Artificial intelligence for TNM staging in NSCLC: a critical appraisal of segmentation utility in [1⁸F]FDG PET/CT. European Journal of Nuclear Medicine and Molecular Imaging. 2026. doi:10.1007/s00259-025-07677-2

BibTeX
@article{Heimer2026,
  author = {Heimer, Maurice M and Dexl, Jakob and Ta, Johanna and Ebner, Ricarda and Herr, Felix L and Orasanin, Leon and Jeblick, Katharina and Adams, Lisa C and Sundar, Lalith K Shiyam and Tufman, Amanda and Werner, Rudolf A and Sheikh, Gabriel and Ricke, Jens and Ingrisch, Michael and Fabritius, Matthias P and Cyran, Clemens C},
  title = {Artificial intelligence for TNM staging in NSCLC: a critical appraisal of segmentation utility in [1⁸F]FDG PET/CT},
  journal = {European Journal of Nuclear Medicine and Molecular Imaging},
  year = {2026},
  doi = {10.1007/s00259-025-07677-2}
}
Kostenlos starten →