ESTRO 2024 - Abstract Book

S1944

Clinical - Mixed sites, palliation

ESTRO 2024

Analysis of 4DMRI visual clarity was performed by 6 independent observers (4 clinical-oncologists and 2 radiologists) who rated optimised versions of 4DMRI AA and 4DMRI CA using a Likert scale, where structures were assigned ratings (‘very clear’ =3, ‘clear = 2, ‘unclear’ = 1 and ‘not visible’ = 0) and assessed for statistical significance using Wilcoxon’s matched-pairs non-parametric test. Composite scores were calculated to provide percentage clarity ratings for all structures. This was done for 3 patients: 1 with locally advanced pancreatic cancer — (LAPC), 1 with a liver metastasis and 1 child post- nephrectomy for Wilms’ tumour. Observers rated confidence in contouring and preference of 4DMRI acquisition plane. Motion analysis was performed by measuring peak-to-peak centre of mass (COM) motion of the dome of diaphragm (DoD), spleen, and left upper renal pole and compared between 4DMRI AA , 4DMRI CA and 4DCT where available. Thirteen patients (Median age = 51 years; range: 5-84 years) underwent 4DMRI (coronal = 13, axial = 9): 5 with LAPC, 4 with liver metastases, and 4 children. It was only possible to obtain 4DCT for 7 patients, 6 had insufficient chest wall motion to capture a trace. Examples of 4DMRI acquired are shown in figure 1. Visual clarity composite scores were greater for 4DMRI AA than 4DMRI CA . For all structures: 56% vs 21% were rated ‘very clear’, 35% vs 60% ‘clear’, 8% vs 18% ‘unclear’, and 1% vs 2% ‘not visible’ for 4DMRI AA vs 4DMRI CA , respectively. Furthermore, for most individual structures, visual clarity ratings were statistically significantly higher for 4DMRI AA than 4DMRI CA . Median score = 3 vs 2 for aorta, liver, kidneys, spinal cord (p = 0.001), median = 3 vs 2 for stomach (p = 0.031), median = 3 vs 2 for pancreas (p=0.016). Only 1 observer preferred 4DMRI CA for the subject with a liver metastasis. All other reviewers favoured 4DMRI AA for the whole cohort. For adult patients, more observers felt ‘very confident’ in using 4DMRI AA than 4DMRI AA (3/5 and 2/5, respectively). For children, 3/5 reviewers did not feel confident in adopting 4DMRI CA , compared with only 1/5 for 4DMRI AA . Results:

There was no significant difference in motion between 4DMRI AA , 4DMRI CA or 4DCT for any structure. Superior inferior (SI) COM motion for all patients for each 4D imaging type is shown in table 1.