[论文解读] Acceleration and Quantitation of Localized Correlated Spectroscopy using Deep Learning: A Pilot Simulation Study

Zohaib Iqbal, Dan Nguyen+2 more|arXiv (Cornell University)|Jun 28, 2018

Advanced MRI Techniques and Applications参考文献 56被引用 4

一句话总结

本研究提出一种基于深度学习的密集连接U-Net（D-UNet）方法，用于加速非均匀采样的局部相关谱学（L-COSY）并准确量化代谢物浓度。该方法在四倍加速下实现低于5%的归一化均方误差，在25%信噪比下误差低于8%，在模拟MRS数据的重建与定量任务中优于压缩感知方法。

ABSTRACT

Nuclear magnetic resonance spectroscopy (MRS) allows for the determination of atomic structures and concentrations of different chemicals in a biochemical sample of interest. MRS is used in vivo clinically to aid in the diagnosis of several pathologies that affect metabolic pathways in the body. Typically, this experiment produces a one dimensional (1D) 1H spectrum containing several peaks that are well associated with biochemicals, or metabolites. However, since many of these peaks overlap, distinguishing chemicals with similar atomic structures becomes much more challenging. One technique capable of overcoming this issue is the localized correlated spectroscopy (L-COSY) experiment, which acquires a second spectral dimension and spreads overlapping signal across this second dimension. Unfortunately, the acquisition of a two dimensional (2D) spectroscopy experiment is extremely time consuming. Furthermore, quantitation of a 2D spectrum is more complex. Recently, artificial intelligence has emerged in the field of medicine as a powerful force capable of diagnosing disease, aiding in treatment, and even predicting treatment outcome. In this study, we utilize deep learning to: 1) accelerate the L-COSY experiment and 2) quantify L-COSY spectra. We demonstrate that our deep learning model greatly outperforms compressed sensing based reconstruction of L-COSY spectra at higher acceleration factors. Specifically, at four-fold acceleration, our method has less than 5% normalized mean squared error, whereas compressed sensing yields 20% normalized mean squared error. We also show that at low SNR (25% noise compared to maximum signal), our deep learning model has less than 8% normalized mean squared error for quantitation of L-COSY spectra. These pilot simulation results appear promising and may help improve the efficiency and accuracy of L-COSY experiments in the future.

研究动机与目标

• 利用深度学习加速耗时的二维L-COSY MRS采集。
• 实现在L-COSY谱中重叠代谢物信号的精确定量。
• 克服传统重建与拟合方法在二维MRS中的局限性，如扫描时间长和光谱拟合复杂。
• 通过模拟代谢物数据验证深度学习在重建非均匀采样L-COSY数据及定量全采样谱方面的性能。
• 通过在真实噪声和采样条件下展示可行性与准确性，为临床转化奠定基础。

提出的方法

• 训练一种密集连接U-Net（D-UNet）架构，从欠采样k空间数据重建非均匀采样（NUS）的L-COSY谱。
• D-UNet模型在使用真实代谢物峰位置、强度和指数线宽展宽生成的模拟L-COSY数据上进行端到端训练。
• 以ℓ₁-范数最小化为基线，对比压缩感知方法的重建性能。
• 对于定量分析，训练另一个独立的D-UNet模型，直接从全采样L-COSY谱预测代谢物浓度。
• 通过在不同信噪比（SNR）和加速因子下计算归一化均方误差（NMSE），评估模型性能。
• 训练数据包含NAA、肌酸、谷氨酸和2-羟基戊二酸（2HG）等代谢物，其峰位置和形状的先验知识已整合到模拟中。

实验结果

研究问题

• RQ1 在高加速因子下，深度学习模型能否以比压缩感知更高的精度重建非均匀采样的L-COSY谱？
• RQ2 在低信噪比条件下，深度学习模型能否在全采样L-COSY谱中准确量化代谢物浓度？
• RQ3 D-UNet模型在模拟L-COSY数据中的重建与定量性能与传统方法相比如何？
• RQ4 该模型在二维MRS中对不同信噪比水平和采样模式的泛化能力如何？
• RQ5 该深度学习框架能否扩展至其他二维MRS技术（如JPRESS或TOCSY）？

主要发现

• 在四倍加速下，D-UNet在重建NUS L-COSY谱时归一化均方误差（NMSE）低于5%，显著优于压缩感知方法（NMSE为20%）。
• 在低信噪比条件（最大信号的25%）下，D-UNet在代谢物定量中保持NMSE低于8%，表现出对噪声的强鲁棒性。
• 与压缩感知相比，D-UNet在重建保真度方面表现更优，尤其在分辨二维谱中重叠的交叉峰方面。
• 该方法在多种信噪比水平和加速因子下均表现出一致性能，表明在模拟数据上具有强大的泛化能力。
• 该框架可扩展至其他二维MRS技术，包括JPRESS、NOESY和TOCSY，如补充结果所示。
• 本研究通过实现采集时间与1D单体素MRS（3–5分钟）相当的L-COSY扫描，突显了其临床转化潜力，使其适用于常规临床应用。