Department of Radiation Oncology, University of California, San Francisco, CA, United States of America.
Siemens Healthineers, Malvern, PA, United States of America.
Phys Med Biol. 2024 Jun 17;69(12). doi: 10.1088/1361-6560/ad5489.
4D MRI with high spatiotemporal resolution is desired for image-guided liver radiotherapy. Acquiring densely sampling k-space data is time-consuming. Accelerated acquisition with sparse samples is desirable but often causes degraded image quality or long reconstruction time. We propose the Reconstruct Paired Conditional Generative Adversarial Network (Re-Con-GAN) to shorten the 4D MRI reconstruction time while maintaining the reconstruction quality.Patients who underwent free-breathing liver 4D MRI were included in the study. Fully- and retrospectively under-sampled data at 3, 6 and 10 times (3×, 6× and 10×) were first reconstructed using the nuFFT algorithm. Re-Con-GAN then trained input and output in pairs. Three types of networks, ResNet9, UNet and reconstruction swin transformer (RST), were explored as generators. PatchGAN was selected as the discriminator. Re-Con-GAN processed the data (3D +) as temporal slices (2D +). A total of 48 patients with 12 332 temporal slices were split into training (37 patients with 10 721 slices) and test (11 patients with 1611 slices). Compressed sensing (CS) reconstruction with spatiotemporal sparsity constraint was used as a benchmark. Reconstructed image quality was further evaluated with a liver gross tumor volume (GTV) localization task using Mask-RCNN trained from a separate 3D static liver MRI dataset (70 patients; 103 GTV contours).Re-Con-GAN consistently achieved comparable/better PSNR, SSIM, and RMSE scores compared to CS/UNet models. The inference time of Re-Con-GAN, UNet and CS are 0.15, 0.16, and 120 s. The GTV detection task showed that Re-Con-GAN and CS, compared to UNet, better improved the dice score (3× Re-Con-GAN 80.98%; 3× CS 80.74%; 3× UNet 79.88%) of unprocessed under-sampled images (3× 69.61%).A generative network with adversarial training is proposed with promising and efficient reconstruction results demonstrated on an in-house dataset. The rapid and qualitative reconstruction of 4D liver MR has the potential to facilitate online adaptive MR-guided radiotherapy for liver cancer.
4D MRI 具有较高的时空分辨率，是图像引导肝脏放射治疗所需要的。获取高密度采样的 k 空间数据是耗时的。使用稀疏采样进行加速采集是可取的，但通常会导致图像质量下降或重建时间延长。我们提出了 Reconstruct Paired Conditional Generative Adversarial Network（Re-Con-GAN），以缩短 4D MRI 重建时间，同时保持重建质量。
本研究纳入了接受自由呼吸肝脏 4D MRI 的患者。首先使用 nuFFT 算法重建全采样和回顾性欠采样数据，倍数分别为 3、6 和 10（3×、6×和 10×）。然后，Re-Con-GAN 对输入和输出进行配对训练。探索了三种类型的网络，ResNet9、UNet 和重建 Swin Transformer（RST），作为生成器。选择 PatchGAN 作为鉴别器。Re-Con-GAN 以时间片（2D +）的形式处理数据（3D +）。共有 48 名患者的 12332 个时间片被分为训练（37 名患者，10721 个切片）和测试（11 名患者，1611 个切片）。使用带有时空稀疏性约束的压缩感知（CS）重建作为基准。使用从单独的 3D 静态肝脏 MRI 数据集（70 名患者；103 个 GTV 轮廓）训练的 Mask-RCNN 进行肝脏大体肿瘤体积（GTV）定位任务，进一步评估重建图像质量。
Re-Con-GAN 的推断时间为 0.15 秒，UNet 和 CS 为 0.16 秒。与 UNet 相比，Re-Con-GAN 和 CS 显著提高了未处理欠采样图像的骰子分数（3× Re-Con-GAN 80.98%；3× CS 80.74%；3× UNet 79.88%）。与 UNet 相比，基于生成对抗网络的训练提出的生成网络具有有前景和高效的重建结果，有望促进在线自适应 MR 引导肝癌放疗。
