ESTRO 2020 Abstract book

S981 ESTRO 2020

PO-1688 Impact of coil selection on tensor-valued dMRI for imaging of the brain using RT fixation P. Brynolfsson 1 , M. Nilsson 2 , L.E. Olsson 1 , F. Szczepankiewicz 2 1 Lund University, Department of Translational Medicine, Lund, Sweden ; 2 Lund University, Dept. of Clinical Sciences, Lund, Sweden Purpose or Objective Using diffusion MRI (dMRI) for treatment response prediction in patients with brain tumors has been an active research field for over 20 years (Ross et al. 1994), so far with limited clinical impact. Unlike conventional dMRI, tensor-valued diffusion encoding can separate effects of microscopic anisotropy, orientation dispersion, and isotropic intravoxel variance with potential applications in response assessment. H&N patients need fixation which prevents the use of a dedicated head coil in favor of flex- coils, with fewer coil elements and suboptimal coverage of the patient anatomy. The aim of this work was to investigate the impact of coil setup on tensor-valued dMRI, adapted to imaging of the brain using RT fixation, with respect to repeatability across coil configurations. Material and Methods A healthy subject was scanned on a GE Discovery 750w 3T with a vendor supplied tensor valued dMRI sequence using TE=124 ms, TR=6364 ms, 2x2x4 mm3 resolution, FOV=240x240x80 mm3, acc=2, scan time 8:42 min, b- values: 0.1, 0.7, 1.4, 2.0 ms/µm2 using 6, 6, 10, 21 linear tensor encoding (LTE) directions. Spherical tensor encoding (STE) was repeated 6, 6, 10, 15 times. A 32- channel head coil and a 6-channel flex coil+posterior array RT setup were compared under repeatability conditions. Tensor-valued dMRI enables estimation of the mean diffusivity (MD), anisotropic kurtosis (MKA) and isotropic kurtosis (MKI) via the signal representation (Westin et al. 2016) : where b is the diffusion weighting, and bΔ is the anisotropy (i.e. shape) of the encoding tensor. STE and LTE correspond to bΔ = 0 and 1, respectively. All data was corrected for Gibbs ringing using subvoxel- shifts. Bland-Altman plots were created to show parameter map agreement between set-ups. Voxelwise SNR was calculated for each b-value from STE measurements to gauge bias due to the noise floor. Signal that follow the Rician distribution will begin to suffer from bias when SNR < 3 (Gudbjartsson et al. 1995). Results Fig.1 shows that parameter maps are qualitatively similar, and Bland-Altman plots of a white matter ROI show quantitative agreement across coil setup with negligible parameter bias. SNR was on average 44% higher on the 32 channel set-up, and was generally above 3 for all b-values apart from the RT coil at the maximal b-value (SNR=3.8±1.4). Prevalence of SNR>3 was 94% and 67% at b = 2.0 ms/µm2, for the 32-ch. and the RT setup, respectively.

PO-1687 Regional lung motion amplitude and variability assessment from a 4DMRI dataset E. Colvill 1 , S. Safai 1 , O. Bieri 2 , S. Kozerke 3 , D.C. Weber 1 , A. Lomax 1 , G. Fattori 1 1 Paul Scherrer Institute, Center for Proton Therapy, Villigen, Switzerland ; 2 University of Basel, Department of Radiology, Basel, Switzerland ; 3 University and ETH Zurich, Institute for Biomedical Engineering, Zurich, Switzerland Purpose or Objective With the aim of modelling the anatomical deformation due to breathing motion, the amplitude and temporal variability have been assessed for different regions in the lungs from the 4DMR images of 14 healthy volunteers and 2,501 breathing cycles from 14 healthy volunteers and two cancer patients have been sampled at 3.33Hz with 4D MR [1] under free-breathing conditions, resulting in a large dataset of 40,099 time-resolved 3D MR volumes. Breathing motion has been modelled using B-spline deformable image registration (DIR) with respect to a reference end- exhale phase, individually selected for each subject. For each case, the specific thoracic geometry has been masked out to limit our analysis at the lungs’ volume, which in turn have been divided into six regions by two sagittal and three axial plane cuts. We report on the mean motion amplitude and its variability, quantified as the interquartile range (IQR) of amplitude, over all the breathing cycles for each mediastinal, lateral and upper, mid, lower regions of each lung. In addition, average breathing period and IQR variation are given from the analysis of the motion dynamic in the right, mid, lateral region. Results The average and range of the mean amplitude (Figure 1a) and variation (Figure 1b) for all voxels within each region of the lungs were calculated. The mean amplitude and variability were greater in the lower regions than the upper ones and in the left lung compared with the right one. Overall, the smallest average motion has been observed in upper-mediastinal region of the right lung with 3.5 mm (IQR: 0.6mm), while the largest values are at lower-lateral side of the left lung, 11.5 mm (IQR: 1.5 mm). The average of the mean breathing period (range) was 4.8(2.4-9.1) seconds with an IQR (range) of 1.3(0.3-3.5) seconds. Conclusion The amplitude and variability of motion in different regions of the lung were assessed using B-spline DIR with differences seen between the left and right lungs as well as within the lungs themselves. The mean breathing period (and variability) also had a broad range from very fast to slow. Understanding the amplitude, breathing period and variability of different portions of the lungs has potential to be used in improving 4D treatment planning and delivery. two cancer patients. Material and Methods

Figure 1. The mean amplitude (a) and variability (b) in 12 lung regions from a dataset of 2,501 breathing cycles of healthy volunteers. Reported values are mean (range) and IQR of the means (range) over the full dataset, respectively for left and right panels. [1] M von Siebenthal, “4D MR imaging of respiratory organ motion and its variability" Phys. Med. Biol 2007

Made with FlippingBook - Online magazine maker