Delineation of Grade II and III Gliomas Investigated by 7T MRI: An Inter-Observer Pilot Study

Martin Prener, Giske Opheim, Helle Juhl Simonsen, Christina Malling Engelmann, Morten Ziebell, Jonathan Carlsen, Olaf B Paulson

Research output: Contribution to journalJournal articleResearchpeer-review

16 Downloads (Pure)

Abstract

PURPOSE: Diffuse low-grade gliomas (DLGGs) are low-malignancy brain tumors originating from the glial cells of the brain growing continuously and infiltratively along the neural axons and infiltrating the surrounding brain tissue. DLGGs usually transform into higher malignancy, causing progressive disability and premature death. MRI scans are valuable when assessing soft tissue abnormalities, but, due to the infiltrative properties of DLGGs, delineating the tumor borders is a challenging task. Therefore, the aim of this study was to explore the difference in gross tumor volume (GTV) of DLGGs delineated from 7 Tesla and 3 Tesla MRI scans.

METHOD: Patients were recruited at the department of neurosurgery and were scanned in both a 7T and a 3T MRI scanner prior to the operation. Two observers delineated the tumors using semi-automatic delineation software. The results from each observer were blinded to the other observer's delineation.

RESULTS: Comparing GTVs from 7T and 3T, the percentage difference varied up to 40.4% on the T2-weighted images. The percentage difference in GTV varied up to 15.3% on the fluid-attenuated inversion recovery (FLAIR) images. On the T2-weighted images, most cases varied by approximately 15%; on the FLAIR sequence, half of the cases varied by approximately 5% and the other half by approximately 15%. The overall inter-observer agreement was near perfect, with an intraclass correlation of 0.969. The intraclass correlation was better on the FLAIR sequence than on the T2 sequence.

CONCLUSION: Overall, the GTVs delineated from 7T images were smaller. The increase in field strength improved the inter-observer agreement only on the FLAIR sequence.

Original languageEnglish
Article number1365
JournalDiagnostics
Volume13
Issue number8
Number of pages10
ISSN2075-4418
DOIs
Publication statusPublished - 7 Apr 2023

Cite this