Multi-Reference Evaluation of Deep Learning Tumour Segmentation in Multi-Sequence Nasopharyngeal Carcinoma MRI
Diagnostics · 03.11.2026 · 16 (21) · 3632
DOI: 10.3390/diagnostics16213632 ↗Kurz gefasst
Fachartikel in der Zeitschrift Diagnostics, online erschienen am 03.11.2026. Verfasst von Ema Pandilova und 5 weiteren Autor:innen. Beteiligte Einrichtung: Faculty of Computer Science and Engineering, Ss. Cyril and Methodius University in Skopje, 1000 Skopje, North Macedonia. Die Arbeit nennt 47 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: European High Performance Computing Joint Undertaking, Ministry of Education and Science, Ss. Cyril and Methodius University in Skopje.
Abstract (Originalsprache)
Background/Objectives: Deep learning segmentation of nasopharyngeal carcinoma is usually reported as one Dice similarity coefficient (DSC) against one reference mask, although multi-sequence MRI is often delineated separately on each sequence. We aimed to quantify how far the choice of reference changes reported performance, and to compare this with the disagreement between radiologists.
Methods
We analysed a public dataset of 277 patient records with separately delineated T1-weighted, T2-weighted and contrast-enhanced T1-weighted MRI, 30 of whom were annotated by two radiologists. Tumour volumes were compared without registration. Spatial agreement was measured after rigid registration, with the loss from transferring masks between acquisition grids estimated separately. nnU-Net models were scored against every reference, including the second radiologist’s, in 102 held-out patient records.
Results
Without registration, the sequence effect on tumour volume was 10.9%, against a reader effect of 5.7% (ratio 1.9, 95% CI 1.40 to 2.49). After rigid registration, per-sequence references agreed at a median DSC of 0.783, against 0.898 between radiologists; in the double-annotated patient records, the paired gap narrowed only from 0.129 to 0.083 (95% CI 0.063 to 0.111) when grid transfer was matched. One fixed set of predictions scored 0.726, 0.685 and 0.673, depending only on the reference (median per-patient reference spread 0.075), and switching the reference sequence moved scores 2.7 to 3.6 times as much as switching the radiologist.
Conclusions
Part of the cross-sequence difference is a difference in what each sequence makes visible rather than annotation error, so a single-reference DSC on multi-sequence data depends on a rarely stated choice. Performance should be reported against every available reference and read against measured inter-reader agreement.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Faculty of Computer Science and Engineering, Ss. Cyril and Methodius University in Skopje, 1000 Skopje, North Macedonia.
ORCID-Profile: Ema Pandilova, Ivan Chorbev, Ivan Kitanovski, Aleksandar Stojmenski, Dimitar Trajanov
Förderung
- European High Performance Computing Joint Undertaking
- Ministry of Education and Science
- Ss. Cyril and Methodius University in Skopje
- Ministry of Education and Science of the Republic of North Macedonia
Zitieren
APA
Pandilova, E., Petrov, M., Chorbev, I., Kitanovski, I., Stojmenski, A., & Trajanov, D. (2026). Multi-Reference Evaluation of Deep Learning Tumour Segmentation in Multi-Sequence Nasopharyngeal Carcinoma MRI. Diagnostics, 16(21), 3632. https://doi.org/10.3390/diagnostics16213632
Vancouver
Pandilova E, Petrov M, Chorbev I, Kitanovski I, Stojmenski A, Trajanov D. Multi-Reference Evaluation of Deep Learning Tumour Segmentation in Multi-Sequence Nasopharyngeal Carcinoma MRI. Diagnostics. 2026;16(21):3632. doi:10.3390/diagnostics16213632
BibTeX
@article{Pandilova2026,
author = {Pandilova, Ema and Petrov, Marko and Chorbev, Ivan and Kitanovski, Ivan and Stojmenski, Aleksandar and Trajanov, Dimitar},
title = {Multi-Reference Evaluation of Deep Learning Tumour Segmentation in Multi-Sequence Nasopharyngeal Carcinoma MRI},
journal = {Diagnostics},
year = {2026},
volume = {16},
number = {21},
pages = {3632},
doi = {10.3390/diagnostics16213632}
}
EUVIMED