3D MRI tumour measurements for treatment planning of eye tumours

Purpose We aim to evaluate 3D definitions of prominence for uveal melanoma (UM), and compare US and contrast‐enhanced T1‐weighted MRI measurements for the most appropriate definitions. Methods 25 UM patients were scanned at 3 Tesla MRI with a 4.7 cm surface coil [1]. The tumour was delineated on the...

Full description

Saved in:
Bibliographic Details
Published in:Acta ophthalmologica (Oxford, England) Vol. 100; no. S267
Main Authors: Klaassen, Lisa, Jaarsma‐Coes, M.G., Ferreira, T.A., Vu, T.H.K., Marinkovic, M., Luyten, G.P.M., Rasch, C.R.N., Beenakker, J.W.M.
Format: Journal Article
Language:English
Published: Malden Wiley Subscription Services, Inc 01-01-2022
Subjects:
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Purpose We aim to evaluate 3D definitions of prominence for uveal melanoma (UM), and compare US and contrast‐enhanced T1‐weighted MRI measurements for the most appropriate definitions. Methods 25 UM patients were scanned at 3 Tesla MRI with a 4.7 cm surface coil [1]. The tumour was delineated on the 3D contrast‐enhanced T1‐weighted images (voxel size 0.8 x 0.8 x 0.8 mm3). Three prominence definitions were evaluated. Promhaus: the distance to the tumour point most distal from sclera. Promsclera: similar to promhaus, but distances are measured perpendicular to the sclera. Promcentre: distance between most interior part of the tumour and sclera. The most appropriate definition was chosen and compared to the clinical ultrasound measurements. Differences >0.5 mm were considered significant. Results For 13/25 patients, there was no significant difference between three definitions. For 12 patients, promsclera or promhaus yielded an oblique measurement or a measurement at the tumour edge. As these measurements are unsuitable for radiotherapy dose calculations, promcentre was selected as the most appropriate definition. For 7 patients, the difference between the US prominence and promcentre was >0.5mm (median absolute difference 0.3 mm, range 0.0–1.6 mm). In 3/7 patients, the MRI measurement was considered more accurate than the US measurement, as the tumour top was not visualized on US. Conclusions After evaluation of 3D prominence definitions for MRI, the median absolute prominence difference (0.3 mm) was below the MRI voxel size and the inter‐observer variability for US measurements [2]. Largest differences were observed in patients where US tumour visualization was incomplete, in which case MRI was preferred. For prominence differences in patients with tumours adjacent to the optic nerve or with enhancing sclera, further study is needed to determine the most accurate method. Bibliography: 1. Ferreira et al. Cancers (2019) 2. Haritoglu et al. Br J Opth (2002)
ISSN:1755-375X
1755-3768
DOI:10.1111/j.1755-3768.2022.173