ESTRO 2021 Abstract Book

S1377

ESTRO 2021

B-value of 800, shows the lowest homogeneity. In terms of SNR, the best averages number for all the considered b-values is 15. The mean ADC value variation on the number of averages is within 3%. When ADC mean values is compared with the reference one obtained with a 1.5T diagnostic MR system, the best b-values choice for the ADC fit was 300-800, if excluded the 0-300 values to exclude any perfusion phenomena. If considered the HMCP, the maximum variation of the ADC mean value due to the averages number among the different Metylcellulose concentration was 0.434 (30 averages, 30%-20%). As shown in figure 2, at least 10 averages should be employed to obtain an accurate ADC value of the evaluated concentrations. The differences in ACD mean values for the five different concentrations of Metycellylose when compared the low and high magnetic field are within 3% and respectively: 30%=0.017, 20%=0.027, 10%= 0.033, 5%=0.034 and 1%=0.041.

Conclusion These preliminary results highlight optimal parameters setting in terms of averages and b-values to achieve a clinical acceptable image quality for the implementation of DWI in the clinical practice. The additional time required to acquire the DWIs is clinically acceptable in our daily workflow. PO-1657 Generation of synthetic CT with 3D deep convolutional neural networks for brain MR-only radiotherapy S. aouadi 1 , R. Hammoud 2 , T. Torfeh 1 , S. Paloor 1 , N. Al-hammadi 1 1 Hamad Medical Corporation, National Centre for Cancer Care and Research, Doha, Qatar; 2 Hamad Medical Corporation, National Centre for Cancer Care and Research, Doha, Qatar Purpose or Objective To create a synthetic CT (sCT) from T2-weighted brain MRI using 3D convolutional neural networks (CNN) algorithm and to assess the resulting image quality in comparison to reference CT. Materials and Methods Conventional T2-weighted MRI (1.5T GE MRI, PROPELLER, TR = 6144.9 ms, TE = 89.82 ms, FA = 160º) and CT datasets from 13 patients who underwent brain radiotherapy were included in this retrospective study. CT and MRI were coregistered and resampled to resolution of 1x1x1mm 3 . The mask of the background was extracted from MRI using the levelsets algorithm. A high-resolution, compact 3D convolutional network was used for the generation of sCT. It used a stack of residual dilated convolutions with increasingly large dilation factors which incorporated large volumetric context. The root mean square error was used as the loss function between sCT and CT. The algorithm is available in the open source NiftyNet library as «highresnet». Geometric assessment of the sCT was performed for all patients using leave one out cross-validation. Voxel- wise Mean Absolute Error (MAE) and Mean Errors (ME) were computed to assess sCT intensities. Bone, soft- tissues and air cavities geometry were quantified by dice (DI), sensitivity(SE) and specificity (SP) indices. MAE Water Equivalent Path Length (MAE_WEPL) was computed for a multitude of spokes starting from the center point of the brain and crossing the whole skull to evaluate the radiologic path length. A comparison with the multiscale and dual contrast patch based method (MDPBM) that was previously published was performed. Results Figure 1 gives the visual assessment of the generated sCT using CNN and shows the average MAE in bins of 20 HU. Mean MAE, ME and MAE_WEPL values for sCT evaluation, using CNN, were 121.92 (σ=21.13), -24.28 (σ=33.73), and 2.18(σ=0.46), respectively. Mean MAE, ME and MAE_WEPL values for sCT evaluation, using MDPBM, were 99.69 (σ=11.07), 2.01 (σ=11.18), and 1.72(σ=0.54), respectively. Table 1 shows DI, SE and SP indices for bone, soft-tissues and air cavities using CNN and MDPBM. MDPM demonstrated better performance than CNN in our dataset.