Extracting diffusion tensor fractional anisotropy and mean diffusivity from 3-direction DWI scans using deep learning
Extracting diffusion tensor fractional anisotropy and mean diffusivity from 3-direction DWI scans using deep learning
复制标题
DOI：
10.1002/mrm.28470
复制
发表时间：
2020-08-18
影响因子：
3.3
通讯作者：
Patel, Sohil H.
中科院分区：
文献类型：
--
作者：
Aliotta, Eric;Nourzadeh, Hamidreza;Patel, Sohil H.
目的开发和评估机器学习方法，从3方向扩散MRI（dMRI）采集中重建分数各向异性（FA）值和平均扩散率（MD）。方法采用两种机器学习模型，分别对低采样dMRI信号和高质量FA和MD图进行映射，FA和MD图是从全采样DTI扫描中重建的。第一个模型是先前描述的多层感知器（MLP），其映射来自单个体素的信号和FA/MD值。第二个是卷积神经网络U-Net模型，它将dMRI切片映射到完整的FA/MD图。每种方法都在dMRI脑扫描（N = 46）上进行训练，并将重建精度与传统的线性最小二乘（LLS）重建进行比较。结果在一个独立的测试队列（N = 20）中，3方向U-Net重建的绝对FA误差显著低于3方向MLP和3方向MLP（U-Net（3-dir）：0.06 +/- 0.01 vs. MLP 3-dir：0.08 +/-0.01，P < 1 x 10（-5））和6方向LLS（LLS 6-dir：0.09 +/-0.03，P = 1 x 10（-5））。3方向MLP（0.06 +/- 0.01 x 10（-3）mm（2）/s）、3方向U-Net（0.06 +/- 0.01 x 10（-3）mm（2）/s）和6方向LLS（0.07 +/- 0.02 x 10（-3）mm（2）/s，P> 0.1）之间的MD误差无显著差异。结论与MLP方法和LLS拟合方法相比，U-Net模型能够更准确地从3个方向的dMRI扫描中重建FA。MD重建精度在重建之间没有显著差异。
Purpose To develop and evaluate machine-learning methods that reconstruct fractional anisotropy (FA) values and mean diffusivities (MD) from 3-direction diffusion MRI (dMRI) acquisitions. Methods Two machine-learning models were implemented to map undersampled dMRI signals with high-quality FA and MD maps that were reconstructed from fully sampled DTI scans. The first model was a previously described multilayer perceptron (MLP), which maps signals and FA/MD values from a single voxel. The second was a convolutional neural network U-Net model, which maps dMRI slices to full FA/MD maps. Each method was trained on dMRI brain scans (N = 46), and reconstruction accuracies were compared with conventional linear-least-squares (LLS) reconstructions. Results In an independent testing cohort (N = 20), 3-direction U-Net reconstructions had significantly lower absolute FA error than both 3-direction MLP (U-Net(3-dir): 0.06 +/- 0.01 vs. MLP3-dir: 0.08 +/- 0.01,P< 1 x 10(-5)) and 6-direction LLS (LLS6-dir: 0.09 +/- 0.03,P= 1 x 10(-5)). The MD errors were not significantly different among 3-direction MLP (0.06 +/- 0.01 x 10(-3)mm(2)/s), 3-direction U-Net (0.06 +/- 0.01 x 10(-3)mm(2)/s), and 6-direction LLS (0.07 +/- 0.02 x 10(-3)mm(2)/s,P> .1). Conclusion The proposed U-Net model reconstructed FA from 3-direction dMRI scans with improved accuracy compared with both a previously described MLP approach and LLS fitting from 6-direction scans. The MD reconstruction accuracies did not differ significantly between reconstructions.