Rachi Toshiya, Tochinai Taku
Department of Radiological Technology, National Cancer Center Hospital East, 6-5-1 Kashiwanoha, Kashiwa City, Chiba, 277-8577, Japan.
Division of Genome and Data Platform, Medical Technology Research and Development Unit, Japan Agency for Medical Research and Development (AMED), 1-7-1 Otemachi, Chiyoda-ku, Tokyo, 100-0004, Japan.
Radiat Oncol. 2025 Aug 4;20(1):122. doi: 10.1186/s13014-025-02708-6.
Proton therapy is commonly used for treating hepatocellular carcinoma (HCC); however, its feasibility can be challenging to assess in large tumors or those adjacent to critical organs at risk (OARs), which are typically assessed only after planning computed tomography (CT) acquisition. This study aimed to predict proton dose distributions using diagnostic CT (dCT) and diagnostic MRI (dMRI) with a convolutional neural network (CNN), enabling early treatment feasibility assessments.
Dose distributions and dose-volume histograms (DVHs) were calculated for 118 patients with HCC using intensity-modulated proton therapy (IMPT) and passive proton therapy. A CPU-based CNN model was used to predict DVHs and 3D dose distributions from diagnostic images. Prediction accuracy was evaluated using mean absolute error (MAE), mean squared error (MSE), peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), and gamma passing rate with a 3 mm/3% criterion.
The predicted DVHs and dose distributions showed high agreement with actual values. MAE remained below 3.0%, with passive techniques achieving 1.2-1.8%. MSE was below 0.004 in all cases. PSNR ranged from 24 to 28 dB, and SSIM exceeded 0.94 in most conditions. Gamma passing rates averaged 82-83% for IMPT and 92-93% for passive techniques. The model achieved comparable accuracy when using dMRI and dCT.
This study demonstrates that early dose distribution prediction from diagnostic imaging is feasible and accurate using a lightweight CNN model. Despite anatomical variability between diagnostic and planning images, this approach provides timely insights into treatment feasibility, potentially supporting insurance pre-authorization, reducing unnecessary imaging, and optimizing clinical workflows for HCC proton therapy.
质子治疗常用于治疗肝细胞癌（HCC）；然而，在大型肿瘤或那些邻近关键危险器官（OARs）的肿瘤中评估其可行性可能具有挑战性，这些肿瘤通常仅在计划计算机断层扫描（CT）采集后进行评估。本研究旨在使用卷积神经网络（CNN）通过诊断CT（dCT）和诊断MRI（dMRI）预测质子剂量分布，从而实现早期治疗可行性评估。
使用调强质子治疗（IMPT）和被动质子治疗计算了118例HCC患者的剂量分布和剂量体积直方图（DVH）。基于CPU的CNN模型用于从诊断图像预测DVH和三维剂量分布。使用平均绝对误差（MAE）、均方误差（MSE）、峰值信噪比（PSNR）、结构相似性指数（SSIM）以及3毫米/3%标准的伽马通过率评估预测准确性。
预测的DVH和剂量分布与实际值高度一致。MAE保持在3.0%以下，被动技术达到1.2 - 1.8%。所有情况下MSE均低于0.004。PSNR范围为24至28分贝，大多数情况下SSIM超过0.94。IMPT的伽马通过率平均为82 - 83%，被动技术为92 - 93%。使用dMRI和dCT时，该模型实现了相当的准确性。
本研究表明，使用轻量级CNN模型从诊断成像进行早期剂量分布预测是可行且准确的。尽管诊断图像和计划图像之间存在解剖学差异，但这种方法能够及时洞察治疗可行性，可能有助于保险预授权、减少不必要的成像，并优化HCC质子治疗的临床工作流程。
