Department of Radiation Oncology, Emory University, Atlanta, Georgia, USA.
School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, Georgia, USA.
Med Phys. 2024 Jun;51(6):4380-4388. doi: 10.1002/mp.17079. Epub 2024 Apr 17.
7 Tesla (7T) apparent diffusion coefficient (ADC) maps derived from diffusion-weighted imaging (DWI) demonstrate improved image quality and spatial resolution over 3 Tesla (3T) ADC maps. However, 7T magnetic resonance imaging (MRI) currently suffers from limited clinical unavailability, higher cost, and increased susceptibility to artifacts.
To address these issues, we propose a hybrid CNN-transformer model to synthesize high-resolution 7T ADC maps from multimodal 3T MRI.
The Vision CNN-Transformer (VCT), composed of both Vision Transformer (ViT) blocks and convolutional layers, is proposed to produce high-resolution synthetic 7T ADC maps from 3T ADC maps and 3T T1-weighted (T1w) MRI. ViT blocks enabled global image context while convolutional layers efficiently captured fine detail. The VCT model was validated on the publicly available Human Connectome Project Young Adult dataset, comprising 3T T1w, 3T DWI, and 7T DWI brain scans. The Diffusion Imaging in Python library was used to compute ADC maps from the DWI scans. A total of 171 patient cases were randomly divided into 130 training cases, 20 validation cases, and 21 test cases. The synthetic ADC maps were evaluated by comparing their similarity to the ground truth volumes with the following metrics: peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), and mean squared error (MSE). In addition, RESULTS: The results are as follows: PSNR: 27.0 ± 0.9 dB, SSIM: 0.945 ± 0.010, and MSE: 2.0E-3 ± 0.4E-3. Both qualitative and quantitative results demonstrate that VCT performs favorably against other state-of-the-art methods. We have introduced various efficiency improvements, including the implementation of flash attention and training on 176×208 resolution images. These enhancements have resulted in the reduction of parameters and training time per epoch by 50% in comparison to ResViT. Specifically, the training time per epoch has been shortened from 7.67 min to 3.86 min.
We propose a novel method to predict high-resolution 7T ADC maps from low-resolution 3T ADC maps and T1w MRI. Our predicted images demonstrate better spatial resolution and contrast compared to 3T MRI and prediction results made by ResViT and pix2pix. These high-quality synthetic 7T MR images could be beneficial for disease diagnosis and intervention, producing higher resolution and conformal contours, and as an intermediate step in generating synthetic CT for radiation therapy, especially when 7T MRI scanners are unavailable.
与 3 特斯拉（3T）ADC 图相比，来自扩散加权成像（DWI）的 7 特斯拉（7T）表观扩散系数（ADC）图具有更高的图像质量和空间分辨率。然而，7T 磁共振成像（MRI）目前存在临床可用性有限、成本更高和对伪影更敏感等问题。
为了解决这些问题，我们提出了一种混合 CNN-Transformer 模型，用于从多模态 3T MRI 中合成高分辨率 7T ADC 图。
提出了 Vision CNN-Transformer（VCT），它由 Vision Transformer（ViT）块和卷积层组成，用于从 3T ADC 图和 3T T1 加权（T1w）MRI 生成高分辨率合成 7T ADC 图。ViT 块能够捕获全局图像上下文，而卷积层则能够有效地捕获细节。该 VCT 模型在公开的 Human Connectome Project Young Adult 数据集上进行了验证，该数据集包含 3T T1w、3T DWI 和 7T DWI 脑扫描。使用 Diffusion Imaging in Python 库从 DWI 扫描中计算 ADC 图。总共 171 个患者病例被随机分为 130 个训练病例、20 个验证病例和 21 个测试病例。通过比较合成 ADC 图与真实体积之间的相似性，使用以下指标评估合成 ADC 图：峰值信噪比（PSNR）、结构相似性指数测量（SSIM）和均方误差（MSE）。此外，结果如下：PSNR：27.0±0.9dB，SSIM：0.945±0.010，MSE：2.0E-3±0.4E-3。定性和定量结果均表明，VCT 表现优于其他最先进的方法。我们引入了各种效率改进，包括实现了 flash attention 和在 176×208 分辨率图像上进行训练。与 ResViT 相比，这些改进使每个时期的参数和训练时间减少了 50%。具体来说，每个时期的训练时间从 7.67 分钟缩短到 3.86 分钟。
我们提出了一种从低分辨率 3T ADC 图和 T1w MRI 预测高分辨率 7T ADC 图的新方法。与 3T MRI 相比，我们预测的图像具有更好的空间分辨率和对比度，并且优于 ResViT 和 pix2pix 的预测结果。这些高质量的合成 7T MRI 图像可能有益于疾病诊断和干预，可以生成更高分辨率和一致的轮廓，并且可以作为生成用于放射治疗的合成 CT 的中间步骤，特别是在没有 7T MRI 扫描仪时。
