快速回波平面磁共振谱成像数据重建算法

传统的相位编码磁共振谱成像采集数据需要很长的时间,使得其在临床上的应用受阻。快速回波平面谱成像(Echoplanarspectroscopicimaging,EPSI)技术采用随时间变化的梯度对谱维和空间维同时进行编码,大大减少了数据采集时间。同时,改进EPSI的读出梯度形式还可以提高‘空间-谱’的分辨率。EPSI数据重建算法比较复杂在t方向先分别对奇偶回波数据进行快速傅立叶变换(FastFouriertranslation,FFT),再利用‘偏移’理论进行组合;在kx方向采用网格化算法将不等间隔采集的数据转换到等间隔的直线网格上,再利用FFT进行图像重建;ky方向是相位编码,不需要转换,直接进行FFT即可。

Data Reconstruction Algorithm on Echo-planar Magnetic Resonance Spectroscopic Imaging

When a conventional phase-encoding magnetic resonance spectrascopic imaging(MRSI) method is in use, the data acquisition is very time consuming and thus it is not well accepted in clinical settings.Echo planar spectroscopic imaging(EPSI) technology has been applied for rapid acquisition of MRSI data. It significantly reduces scan time by simultaneously encoding the spectral dimension and the spatial dimension. Reconstruction of EPSI data is sophisticated. For the t-dimension, data are non-uniformly sampled. Fast Fourier transform(FFT) is applied in odd echoes and even echoes respectively and then shift method is used to combine them together. For the k_x dimension, the data are unevenly sampled and have to be interpolated onto a Cartesian grid before FFT is applied. For the phase-encoding dimension (k_y), the Fourier transform can be readily done with FFT since the data are evenly spaced.