ESTRO 2022 - Abstract Book

S50

Abstract book

ESTRO 2022

Conclusion By combining a navigator-triggered MVXD image with an SMS-4D-MRI image, we constructed a dramatically sharper midP image. DDM analysis shows that the MVXD-midP is slightly less reliable than the 4D-midP, but this has limited influence on the regions of interest. Using this higher-resolution MVXD-midP could improve treatment planning.

PD-0075 3D MRI versus conventional ultrasound tumour measurements for treatment planning of eye tumours

L. Klaassen 1,2,3 , M. Jaarsma-Coes 2,1 , T. Ferreira 2 , K. Vu 1 , M. Marinkovic 4 , G. Luyten 1 , C. Rasch 3 , J. Beenakker 1,2,5

1 Leiden University Medical Center, Ophthalmology, Leiden, The Netherlands; 2 Leiden University Medical Center, Radiology, Leiden, The Netherlands; 3 Leiden University Medical Center, Radiation Oncology, Leiden, The Netherlands; 4 Leiden University Medical Center, Ophthalmology, Leide, The Netherlands; 5 Leiden University Medical Center, Radiation Oncology, Leiden, The Netherlands Purpose or Objective Uveal melanoma (UM) is the most frequently occurring malignant primary eye tumour. Ocular MRI, enabling 3D tumour imaging and providing a better soft tissue contrast than conventional 2D ultrasound (US), is increasingly used for the diagnosis, therapy planning and follow-up of UM. Tumour prominence (thickness) and largest basal diameter (LBD) are the primary determinants for the brachytherapy applicator size and application time and are used to define the 3D clinical target volume in proton beam therapy. Within this study, we aim to compare US and MRI dimension measurements and determine the most suitable measurement modality. Materials and Methods Data of 25 UM patients with a wide range in size and location were analysed retrospectively. Patients were scanned at 3 Tesla MRI according to the protocol of Ferreira (Cancers 2019). The tumour was semi-automatically delineated on the contrast-enhanced T1-weighted images (acquisition voxel size 0.8x0.8x0.8 mm 3 ) in MevisLab. The prominence and LBD were automatically obtained from the MRI contours. MRI measurements were compared to the clinical US measurements (resolution 0.4 mm) and related to tumour location in the anteroposterior direction. Tumours where the measurement did not fit into the US field of view (FOV) or where the extent of flat tumour components was difficult to assess on MRI were assessed separately. Results For 7/25 patients, the prominence and/or LBD measurement did not fit into the US FOV. All these tumours were located in the anterior 50% of the eye. Of the 4 patients with an US prominence <4 mm, tumour extent was difficult to assess on MRI for 3/4 (Fig 1). When tumour extent was visible on both imaging modalities, median absolute differences were 0.3 mm (range 0.0-1.3 mm) for prominence and 1.1 mm (range 0.1-2.8 mm) for LBD, respectively (Fig 2). For the LBD, 80% of measurements was more than 0.5 mm larger on MRI. For patients for whom the tumour extent was difficult to assess on one imaging modality, the median absolute difference for the prominence and LBD were 0.7 mm (range 0.5-1.6 mm) and 1.4 mm (range 0.1-7.6 mm).