ESTRO 2024 - Abstract Book

S4099

Physics - Inter-fraction motion management and offline adaptive radiotherapy

ESTRO 2024

2631

Digital Poster

Multi-observer investigation of uncertainties in MRI-only prostate fiducial marker identification

Emilia Persson 1,2 , Sacha Af Wetterstedt 1 , Jonas Scherman 1 , Christian Jamtheim Gustafsson 1,2 , Per Munck af Rosenschold 1,3 , Lars E Olsson 1,2 , André Haraldsson 1,3 1 Radiation Physics, Department of Hematology, Oncology, and Radiation Physics, Skåne University Hospital, Lund, Sweden. 2 Medical Radiation Physics, Department of Translational Medicine, Lund University, Malmö, Sweden. 3 Medical Radiation Physics, Department of Clinical Sciences, Lund University, Lund, Sweden

Purpose/Objective:

Motivated by reduced uncertainties and workflow optimization, magnetic resonance imaging (MRI)-only radiotherapy is today used in clinical practice worldwide. For prostate cancer patients, commonly positioned using implanted fiducial markers, new strategies for marker identification and reference image creation are required. A common strategy is to identify the center positions of the markers in the MR-images and use this as a reference throughout the workflow. Since the center position can be defined in multiple ways, and is likely to vary between observers, the marker identification process introduces spatial uncertainties into the workflow. The aim of this study was to evaluate the uncertainty in fiducial marker definition in MRI-only radiotherapy of prostate cancer due to the impact of observer variability and the limitation that comes from marker definition limited to a discrete MR-image slice. By comparing the marker locations of multiple observer identifications to the actual clinical marker locations, the combined effect of these factors was investigated. Nine patients included in the REMIND study (NCT05844761) were retrospectively investigated in this study. All patients received hypofractionated treatment of the prostate gland with 42.7Gy/7 fractions. Patient positioning was based on three implanted cylindrical gold fiducial markers with the dimensions 1x5mm. The center positions of the marker were identified using MR-images in a commercially available software. A transversal large field-of-view T2 weighted planning sequence with 2.5 mm slice thickness and a multi-echo gradient-echo sequence were used to help facilitate the marker identification. According to clinical practice, the identified center position of the marker was limited to be identified in one discrete slice of the MR-images, and not in between slices. Four observers retrospectively identified the implanted fiducial markers using the same method as above but without restricting the center position to a discrete slice in the MR-images. All observers were provided the same instructions and had previous experience of marker identification in MR-images. Identification was performed individually by each observer without the knowledge of the other observers or clinical marker positions. The new marker positions were compared to the marker positions used clinically by calculation of the absolute difference between markers. Material/Methods:

Results: