LAUSR

BACKGROUND Despite the significant physical differences between MRI and CT, the high entropy of MRI data indicates the existence of a surjective transformation from MRI to CT image. However, there is no specific optimization of the network itself in previous MRI/CT translation works, resulting in mistakes in details such as the skull margin and cavity edge. These errors might have moderate effect on conventional radiotherapy, but for BNCT, the skin dose will be a critical part of the dose composition. PURPOSE To create a self-attention network which could directly transfer magnetic resonance imaging (MRI) to synthetical computerized tomography (sCT) images with lower inaccuracy at the skin edge and examine the viability of MR-guided BNCT. METHODS A retrospective analysis was undertaken on 104 patients with brain malignancies who had both CT and MRI as part of their radiation treatment plan. The CT images were deformably registered to the MRI. In the U-shaped generation network, we introduced spatial and channel attention modules, as well as a versatile "Attentional ResBlock", which reduces the parameters while maintaining high performance. We employed 5-fold cross-validation to test all patients, compared the proposed network to those used in earlier studies and used Monte Carlo software to simulate the BNCT process for dosimetric evaluation in test set. RESULTS Compared with Unet, Pix2Pix and ResNet, the mean absolute error (MAE) of SARU is reduced by 12.91 HU, 17.48 HU and 9.50 HU respectively. The " Two One-Sided Test" shows no significant difference in dose-volume histogram (DVH) results. And for all tested cases, the average 2%/2mm gamma index of Unet, ResNet, Pix2Pix and SARU were 0.96 ± 0.03, 0.96 ± 0.03, 0.95 ± 0.03 and 0.98 ± 0.01, respectively. The error of skin dose from SARU is much less than the results from other methods. CONCLUSION We have developed a residual U-shape network with an attention mechanism to generate sCT images from MRI for BNCT treatment planning with lower MAE in six organs. There is no significant difference between the dose distribution calculated by sCT and real CT. This solution may greatly simplify the BNCT treatment planning process, lower the BNCT treatment dose, and minimize image feature mismatch. This article is protected by copyright. All rights reserved.