Adaptive distortion correction techniques for high-field magnetic resonance neuroimaging

近年来，单次激发回波平面成像(EPI)的使用有了显著增长。在功能神经成像中，许多先进的脑成像应用，如功能磁共振成像(FMRI)和扩散张量成像(DTI)，都使用EPI作为成像模块。然而，由于样品内主磁场的不均匀，EPI容易受到许多成像伪影和几何图像失真的影响。这些由样品中的磁化率差异或主场不均匀性引起的不均匀性通常被认为是时间上的固有常数。特别是对于具有连续和重复扫描、系统不稳定、铁垫片加热或受试者S运动和呼吸的长时间序列采集，可能会在场的均匀性中引入随时间的变化。这些结果导致图像失真的时变变化，导致退化和时间序列分析中的伪影。在3特斯拉以上的高场强下，这些效应变得越来越重要，并可能成为7T时EPI应用的一个重要障碍。在这个项目中，我们提出了一种动态测量、表征和补偿EPI中几何图像失真的方法。该项目涵盖了从功能磁共振到扩散张量成像的各种EPI应用。这些即将开发的方法将在1.5T到7T的不同场强下进行评估，它们将显著改善EPI的稳定性和成像性能，以实现更灵敏的实验结果和更高的定位精度。

The use of single-shot echo-planar imaging (EPI) has grown dramatically over recent years. A number of advanced brain imaging applications in functional neuroimaging, e.g. functional MRI (fMRI) and diffusion tensor imaging (DTI), use EPI as the imaging module. EPI, however, is susceptible to a number of imaging artifacts and geometric image distortions due to inhomogeneities of the main magnetic field within the sample. These inhomogeneities induced by magnetic susceptibility differences in the sample or by main field inhomogeneities are often regarded as inherently constant in time. Especially for long time-series acquisitions with continuous and repetitive scans, system instabilities, ferro-shim heating or subject¿s motion and respiration can introduce time dependant changes in the field homogeneity. These result in time variant changes in image distortions leading to degradation and to artifacts in time-series analyses. These effects become increasingly important at high field strengths of 3 Tesla and above, and can become a significant obstacle for EPI applications at 7 T. In this project we propose an approach to dynamically measure, characterise and compensate for geometric image distortions in EPI. The project covers various EPI applications ranging from functional MRI to diffusion tensor imaging. The methods, which are to be developed, will be evaluated at various field strengths ranging from 1.5 T to 7 T. They will significantly improve stability and imaging properties of EPI to allow more sensitive experimental results and higher positional accuracy.