Generating Synthetic Computed Tomography and Synthetic Magnetic Resonance (sMR: sT1w/sT2w) Images of the Brain Using Atlas-Based Method

Abstract

<strong><em>Introduction:</em></strong> Nowadays, magnetic resonance imaging (MRI) in combination with computed-tomography (CT) is increasingly being used in radiation therapy planning. MR and CT images are applied to determine the target volume and calculate dose distribution, respectively. Since the use of these two imaging modalities causes registration uncertainty and increases department workload and costs, in this study, brain synthetic CT (sCT) and synthetic MR (sMR: sT1w/sT2w) images were generated using Atlas-based method; consequently, just one type of image (CT or MR) is taken from the patient.<br /> <strong><em>Material and Methods:</em></strong> The dataset included MR and CT paired images from 10 brain radiotherapy (RT) patients. To generate sCT/sMR images, first each MR/CT Atlas was registered to the MR/CT target image, the resulting transformation was applied to the corresponding CT/MR Atlas, which created the set of deformed images. Then, the deformed images were fused to generate a single sCT/sMR image, and finally, the sCT/sMR images were compared to the real CT/MR images using the mean absolute error (MAE).<br /> <strong><em>Results:</em></strong> The results showed that the MAE of sMR (sT1w/sT2w) was less than that of sCT images. Moreover, sCT images based on T1w were in better agreement with real CT than sCT-based T2w. In addition, sT1w images represented a lower MAE relative to sT2w.<br /> <strong><em>Conclusion: </em></strong>The CT target image was more successful in transferring the geometry of the brain tissues to the synthetic image than MR target.