Department of Biomedical Engineering, Columbia University, New York, New York, USA.
Department of Psychiatry, Columbia University, New York, New York, USA.
J Magn Reson Imaging. 2024 Mar;59(3):964-975. doi: 10.1002/jmri.28868. Epub 2023 Jul 4.
Deep learning-based methods have been successfully applied to MRI image registration. However, there is a lack of deep learning-based registration methods for magnetic resonance spectroscopy (MRS) spectral registration (SR).
To investigate a convolutional neural network-based SR (CNN-SR) approach for simultaneous frequency-and-phase correction (FPC) of single-voxel Meshcher-Garwood point-resolved spectroscopy (MEGA-PRESS) MRS data.
Retrospective.
Forty thousand simulated MEGA-PRESS datasets generated from FID Appliance (FID-A) were used and split into the following: 32,000/4000/4000 for training/validation/testing. A 101 MEGA-PRESS medial parietal lobe data retrieved from the Big GABA were used as the in vivo datasets.
FIELD STRENGTH/SEQUENCE: 3T, MEGA-PRESS.
Evaluation of frequency and phase offsets mean absolute errors were performed for the simulation dataset. Evaluation of the choline interval variance was performed for the in vivo dataset. The magnitudes of the offsets introduced were -20 to 20 Hz and -90° to 90° and were uniformly distributed for the simulation dataset at different signal-to-noise ratio (SNR) levels. For the in vivo dataset, different additional magnitudes of offsets were introduced: small offsets (0-5 Hz; 0-20°), medium offsets (5-10 Hz; 20-45°), and large offsets (10-20 Hz; 45-90°).
Two-tailed paired t-tests for model performances in the simulation and in vivo datasets were used and a P-value <0.05 was considered statistically significant.
CNN-SR model was capable of correcting frequency offsets (0.014 ± 0.010 Hz at SNR 20 and 0.058 ± 0.050 Hz at SNR 2.5 with line broadening) and phase offsets (0.104 ± 0.076° at SNR 20 and 0.416 ± 0.317° at SNR 2.5 with line broadening). Using in vivo datasets, CNN-SR achieved the best performance without (0.000055 ± 0.000054) and with different magnitudes of additional frequency and phase offsets (i.e., 0.000062 ± 0.000068 at small, -0.000033 ± 0.000023 at medium, 0.000067 ± 0.000102 at large) applied.
The proposed CNN-SR method is an efficient and accurate approach for simultaneous FPC of single-voxel MEGA-PRESS MRS data.
4 TECHNICAL EFFICACY: Stage 2.
深度学习方法已成功应用于磁共振成像（MRI）图像配准。然而，在磁共振波谱（MRS）谱配准（SR）方面，缺乏基于深度学习的配准方法。
研究一种基于卷积神经网络的磁共振波谱（MRS）谱配准（CNN-SR）方法，用于单像素 Meshcher-Garwood 点分辨波谱（MEGA-PRESS）MRS 数据的频率和相位同时校正（FPC）。
回顾性。
使用来自 FID Appliance（FID-A）的 40000 个模拟 MEGA-PRESS 数据集，并将其分为以下部分：32000/4000/4000 用于训练/验证/测试。从 Big GABA 中检索到的 101 个 MEGA-PRESS 中脑叶数据被用作体内数据集。
场强/序列：3T，MEGA-PRESS。
对模拟数据集的频率和相位偏移均方根误差进行评估。对体内数据集的胆碱间隔方差进行评估。在不同信噪比（SNR）水平下，模拟数据集的偏移量大小为-20 至 20 Hz 和-90°至 90°，分布均匀。对于体内数据集，引入了不同大小的额外偏移量：小偏移量（0-5 Hz；0-20°）、中偏移量（5-10 Hz；20-45°）和大偏移量（10-20 Hz；45-90°）。
使用模拟和体内数据集的模型性能的双侧配对 t 检验，P 值<0.05 被认为具有统计学意义。
CNN-SR 模型能够校正频率偏移（SNR 为 20 时为 0.014±0.010 Hz，SNR 为 2.5 时为 0.058±0.050 Hz，线展宽）和相位偏移（SNR 为 20 时为 0.104±0.076°，SNR 为 2.5 时为 0.416±0.317°，线展宽）。使用体内数据集，CNN-SR 在没有（0.000055±0.000054）和有不同大小的额外频率和相位偏移（即小偏移量为 0.000062±0.000068，中偏移量为-0.000033±0.000023，大偏移量为 0.000067±0.000102）时都取得了最佳性能。
所提出的 CNN-SR 方法是一种用于单像素 MEGA-PRESS MRS 数据频率和相位同时校正的有效且准确的方法。
4 级 技术功效：2 级。
