ESTRO 2021 Abstract Book

S144

ESTRO 2021

gross tumour volume (GTV) regions, and fed into simulations to compare b-opt₇ to b-im (Wilcoxon tests, α=0.05) and characterize tissue perfusion regimes. Results Lower σ ₜₒₜ were obtained for b-opt₁₃ and b-opt₇ with respect to b-im for Medium and High regimes (Figure 1), suggesting that clinical protocols would benefit from an optimization if the expected perfusion regimes were high enough. As expected, σ ₜₒₜ was lower for b-opt₁₃ and estimates significantly differed in Medium and High regimes with respect to b-im and b-opt₇, showing that simplified protocols imply significantly different estimates. As for imaging biomarkers estimation, IVIM parameters from the GTV were comparable to Medium Perfusion, whereas those from WM to Low Perfusion, in both chordoma and meningioma patients (Figure 2).

Conclusion The proposed framework showed that b-values optimization can aid the estimation of IVIM parameters to derive accurate brain tumours biomarkers with potential applications in personalized oncology. Nevertheless in-vivo IVIM measurements with optimal b-values are needed to validate these findings.

PH-0213 Repeatability of free breathing diffusion weighted MRI for MR guided liver SBRT V. van Pelt 1 , R.J.M. Navest 1 , M.E. Nowee 1 , U.A. van der Heide 1 , P.J. van Houdt 1 1 The Netherlands Cancer Institute, Radiation Oncology, Amsterdam, The Netherlands

Purpose or Objective Online adaptive MR-guided Stereotactic Body Radiotherapy (SBRT) of liver metastases is an upcoming approach 1 . Daily functional imaging on a MR-linac provides the opportunity to measure tumor biology changes during treatment. The apparent diffusion coefficient (ADC) derived from diffusion weighted MRI (DWI) is a potential candidate to monitor such changes. Key for success of a response biomarker is that the measurements are accurate and precise over time 2 . Therefore, the aim of this study was to assess the repeatability of ADC values from free-breathing DWI measured on an MR-linac. Materials and Methods 18 patients were treated with 3x20Gy on a 1.5T MR-linac (Unity, Elekta AB, Stockholm, Sweden) in a prospective trial. DWI scans were performed in 2 separate sessions (test-retest) before treatment delivery (free breathing, b-values 0, 30, 150, and 500s/mm 2 , FOV=510x319x98mm 3 , voxel size= 4x4x4mm 3 ). The ADC was calculated from the high b-value images (150 and 500 s/mm 2 ). The tumor was delineated on the b= 500s/mm 2 by two observers. The mean of ADC values were obtained for each tumor delineation. The test- retest scans were used to determine the repeatability coefficient (%RC) from the within-subject coefficient of variation (wCV) 2 . The inter-observer variation was assessed by wCV of the ADC values from the two delineations on the test scans. To determine whether the repeatability is influenced by the amount of tumor motion, the motion amplitude from the 4D T2-weigthed MRI was related to the percentage difference in ADC of the test-retest data with Pearson correlation coefficient. Results Fig. 1 shows an example of the test-retest data including delineations. The size of the tumors and mean ADC values ranged between 4.0-80.0 cm 3 and 0.4-1.7 mm 2 /s. The %RC of the ADC values was 41% (wCV 15%) and 45% (wCV 16%) for observer 1 and observer 2, respectively. The inter-observer wCV for ADC was 3.4%. The