Zou Chao, Tie Changjun, Pan Min, Wan Qian, Liang Changhong, Liu Xin, Chung Yiu-Cho
Paul C Lauterbur Research Center for Biomedical Imaging, Shenzhen Key Laboratory for MRI, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong, People's Republic of China.
Phys Med Biol. 2017 Jan 7;62(1):1-16. doi: 10.1088/1361-6560/62/1/1. Epub 2016 Dec 14.
Proton resonance frequency shift (PRFS) MR thermometry is commonly used to measure temperature in thermotherapy. The method requires a baseline temperature map and is therefore motion sensitive. Several referenceless MR thermometry methods were proposed to address this problem but their performances have never been compared. This study compared the performance of five referenceless methods through simulation, heating of ex vivo tissues and in vivo imaging of the brain and liver of healthy volunteers. Mean, standard deviation, root mean square, 2/98 percentiles of error were used as performance metrics. Probability density functions (PDF) of the error distribution for these methods in the different tests were also compared. The results showed that the phase gradient method (PG) exhibited largest error in all scenarios. The original method (ORG) and the complex field estimation method (CFE) had similar performance in all experiments. The phase finite difference method (PFD) and the near harmonic method (NH) were better than other methods, especially in the lower signal-to-noise ratio (SNR) and fast changing field cases. Except for PG, the PDFs of each method were very similar among the different experiments. Since phase unwrapping in ORG and NH is computationally demanding and subject to image SNR, PFD and CFE would be good choices as they do not need phase unwrapping. The results here would facilitate the choice of appropriate referenceless methods in various MR thermometry applications.
质子共振频率偏移（PRFS）磁共振测温法常用于热疗中的温度测量。该方法需要一个基线温度图，因此对运动敏感。为解决这一问题，人们提出了几种无参考磁共振测温法，但从未对它们的性能进行过比较。本研究通过模拟、离体组织加热以及对健康志愿者的大脑和肝脏进行体内成像，比较了五种无参考方法的性能。使用误差的均值、标准差、均方根、2/98百分位数作为性能指标。还比较了这些方法在不同测试中误差分布的概率密度函数（PDF）。结果表明，相位梯度法（PG）在所有情况下误差最大。原始方法（ORG）和复场估计法（CFE）在所有实验中表现相似。相位有限差分法（PFD）和近谐波法（NH）优于其他方法，尤其是在低信噪比（SNR）和快速变化场的情况下。除PG外，各方法的PDF在不同实验中非常相似。由于ORG和NH中的相位展开计算量较大且受图像SNR影响，PFD和CFE是不错的选择，因为它们不需要相位展开。本文结果将有助于在各种磁共振测温应用中选择合适的无参考方法。
