Contrast-Enhanced Liver Magnetic Resonance Image Synthesis Using Gradient Regularized Multi-Modal Multi-Discrimination Sparse Attention Fusion GAN.
Contrast-Enhanced Liver Magnetic Resonance Image Synthesis Using Gradient Regularized Multi-Modal Multi-Discrimination Sparse Attention Fusion GAN.
复制标题
DOI：
10.3390/cancers15143544
复制
发表时间：
2023-07-08
期刊：
影响因子：
5.2
通讯作者：
Yang, Wensha
中科院分区：
文献类型：
--
作者：
Jiao, Changzhe;Ling, Diane;Bian, Shelly;Vassantachart, April;Cheng, Karen;Mehta, Shahil;Lock, Derrick;Zhu, Zhenyu;Feng, Mary;Thomas, Horatio;Scholey, Jessica E.;Sheng, Ke;Fan, Zhaoyang;Yang, Wensha
对比增强MR已被用于肝脏患者的诊断和治疗。最近，MR引导的放射治疗的发展要求每日对比MR用于肿瘤靶向。然而，频繁注射造影剂对患者有风险。我们开发了一个深度学习模型（GRMM-GAN），用于从造影前图像合成对比增强MR。GRMM-GAN采用梯度正则化和多重判别机制。与最先进的深度学习模型相比，它显示出上级性能。用途：为了提供腹部对比增强MR图像合成，我们开发了一种梯度正则化多模态多分辨稀疏注意力融合生成对抗网络（GRMM-GAN），以避免对患者重复注射对比剂并促进自适应监测。方法：经IRB批准，从我们的机构数据库中回顾性征集了61例肝癌患者的165项腹部MR研究。每项研究均包括T2、T1造影前（T1 pre）和T1造影增强（T1 ce）图像。GRMM-GAN合成管道由稀疏注意力融合网络，图像梯度正则化器（GR）和具有多分辨力的生成对抗网络组成。这些研究被随机分为115项用于培训，20项用于验证，30项用于测试。在训练阶段，采用两种对比前MR模式（T2和T1 pre图像）作为输入。门静脉期的T1 ce图像用作输出。将合成的T1 ce图像与地面真实T1 ce图像进行比较。评价指标包括峰值信噪比（PSNR），结构相似性指数（SSIM）和均方误差（MSE）。图灵测试和专家的轮廓评价图像合成质量。结果：所提出的GRMM-GAN模型的PSNR为28.56，SSIM为0.869，MSE为83.27。与最先进的模型比较相比，所提出的模型在所有测试指标中均显示出统计学显著性改善，p值< 0.05。平均图灵测试得分为52.33%，接近随机猜测，支持该模型的临床应用的有效性。在肿瘤特异性区域分析中，合成MR图像的平均肿瘤对比噪声比（CNR）与真实的MR图像相比无统计学显著性。来自真实的与合成图像的平均DICE为0.90，而操作员间DICE为0.91。结论：我们展示了一种新的多模态MR图像合成神经网络GRMM-GAN的功能，用于基于预对比T1和T2 MR图像的T1 ce MR合成。GRMM-GAN有望避免放射治疗期间的重复造影剂注射。
Contrast-enhanced MR has been used in diagnosing and treating liver patients. Recently, development in MR-guided radiation therapy calls for daily contrast MR for tumor targeting. However, frequent contrast injection is risky to patients. We developed a deep learning model (GRMM-GAN) to synthesize contrast-enhanced MR from pre-contrast images. GRMM-GAN adopts gradient regularization and multi-discrimination mechanisms. It shows superior performance compared with state-of-the-art deep learning models. Purposes: To provide abdominal contrast-enhanced MR image synthesis, we developed an gradient regularized multi-modal multi-discrimination sparse attention fusion generative adversarial network (GRMM-GAN) to avoid repeated contrast injections to patients and facilitate adaptive monitoring. Methods: With IRB approval, 165 abdominal MR studies from 61 liver cancer patients were retrospectively solicited from our institutional database. Each study included T2, T1 pre-contrast (T1pre), and T1 contrast-enhanced (T1ce) images. The GRMM-GAN synthesis pipeline consists of a sparse attention fusion network, an image gradient regularizer (GR), and a generative adversarial network with multi-discrimination. The studies were randomly divided into 115 for training, 20 for validation, and 30 for testing. The two pre-contrast MR modalities, T2 and T1pre images, were adopted as inputs in the training phase. The T1ce image at the portal venous phase was used as an output. The synthesized T1ce images were compared with the ground truth T1ce images. The evaluation metrics include peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), and mean squared error (MSE). A Turing test and experts’ contours evaluated the image synthesis quality. Results: The proposed GRMM-GAN model achieved a PSNR of 28.56, an SSIM of 0.869, and an MSE of 83.27. The proposed model showed statistically significant improvements in all metrics tested with p-values < 0.05 over the state-of-the-art model comparisons. The average Turing test score was 52.33%, which is close to random guessing, supporting the model’s effectiveness for clinical application. In the tumor-specific region analysis, the average tumor contrast-to-noise ratio (CNR) of the synthesized MR images was not statistically significant from the real MR images. The average DICE from real vs. synthetic images was 0.90 compared to the inter-operator DICE of 0.91. Conclusion: We demonstrated the function of a novel multi-modal MR image synthesis neural network GRMM-GAN for T1ce MR synthesis based on pre-contrast T1 and T2 MR images. GRMM-GAN shows promise for avoiding repeated contrast injections during radiation therapy treatment.
登录
查看更多内容
DOI：
10.1023/b:visi.0000029664.99615.94
发表时间：
2004-11-01
影响因子：
19.5
作者：
Lowe, DG
通讯作者：
Lowe, DG
影响因子：
4.4
作者：
Montalt-Tordera, Javier;Quail, Michael;Steeden, Jennifer A.;Muthurangu, Vivek
通讯作者：
Muthurangu, Vivek
用于低剂量 CT 图像增强的梯度正则化卷积神经网络
影响因子：
3.5
作者：
Gou, Shuiping;Liu, Wei;Jiao, Licheng
通讯作者：
Jiao, Licheng
影响因子：
5.2
作者：
通讯作者：
--
影响因子：
45.3
作者：
Gordan, John D.;Kennedy, Erin B.;Rose, Michal G.
通讯作者：
Rose, Michal G.