ESTRO 2023 - Abstract Book

S1375

Digital Posters

ESTRO 2023

Generative Adversarial Network (cycleGAN) combines two GAN in opposition and has a ResNet 9 block as a generator and a 70*70*70 PatchGAN network as discriminator. The images were distributed into training cohorts of 20 patients and test cohorts of 10 patients. A cohort of 10 patients from each center was used for training in the multicenter study. All the images in the train were cut into four patches that overlapped on the half and data augmentation through rotations is applied during training with a maximum angle of 30°. The loss functions used are the L1 norm as cycle loss with a weight of 30, a binary cross-entropy (bce) as adversarial and a L1 norm as identity, both with a weight of 1. The learning rate is worth 0.0001 and there are 100 epochs. Synthetic CTs (sCTs) were compared to the original CTs by a voxel-wise comparison on body and bones with the mean absolute error (MAE) in Hounsfield units (HU), mean error (ME) in HU too and peak signal to noise ratio (PSNR) in dB. The Wilcoxon test was used to compare the results obtained with the different trainings. Significant differences were considered for p-value < 0.05. Results Table 1 shows MAE, ME and PSNR for body and bone contours for monocenter and multicenter training cohort. Figure 1 presents a comparison of the sCT obtained with the different tests. There is a presence of artifacts during one train from center 2 followed by a test from center 1. The Wilcoxon test shows a significant difference between monocentric studies and those with a different training and testing center. On the contrary, the p-value is high when comparing the monocenter study with the study that has a composite train of data from both centers.

Conclusion This comparison between the monocenter and multicenter trainings for two different centers showed that there was no significant difference in terms of outcomes and images when the learning cohort contains data from both centers.

PO-1672 A pre-processing pipeline for 0.35T pelvic MR image quality enhancement: development and evaluation

M. Vagni 1 , H.E. Tran 1 , L. Boldrini 1 , G. Chiloiro 1 , A. D'Aviero 2 , A. Re 2 , A. Romano 1 , L. Indovina 1 , V. Valentini 1 , D. Cusumano 1 , L. Placidi 1 1 Fondazione Policlinico Universitario "A. Gemelli" IRCCS, Department of Radiation Oncology, Rome, Italy; 2 Mater Olbia Hospital, Department of Radiation Oncology, Olbia (SS), Italy Purpose or Objective MR-Linac systems offer high soft-tissue contrast volumes thanks to the onboard MR-scanner. However, low frequency intensity inhomogeneity artefacts can occur, resulting in signal losses that may affect the image quality (IQ). Various correction strategies have been proposed to reduce such artefacts on high-field MRIs and very limited cases are focused on 0.35T ones. This study aims to outline and assess an image processing pipeline able to enhance the IQ, offering a better visualization of targets and organs at risk (OARs) and a more reliable intensity distribution for the subsequent digital processing, such as auto-segmentation. Materials and Methods 0.35T simulation MRIs (ORIG) from 73 prostate cancer patients treated with MR-Linac were collected. The N4 bias field correction (N4ITK), alone or in conjunction with an Adaptive Histogram Equalization filter (N4ITK_AHE), was applied to the images. Four radiation oncologists with different experience (O1&O2: more than 5 years of experience; O3&O4: less than