RIA: Efficient Dynamic Magnetic Resonance Imaging with A Priori Constraints

随着功能磁共振成像为研究人脑功能提供了一种新的方法，提高实验的时间分辨率以捕捉快速的瞬时神经元事件变得越来越重要。在传统的基于傅里叶变换的成像方法中，每个动态图像都是独立获取的，因此可能的时间分辨率受到在傅立叶空间采集的数据点数量的限制，从而达到一定的空间分辨率水平。这项研究试图通过将先验信息融入到成像过程中来克服这个问题。具体地说，正在开发用于数据采集和图像重建的新的优化方法，其利用信号的基本模型以及关于被成像的对象和动态变化的可用的先验信息。这些方法有望在成像效率上比传统的傅立叶成像方法提高一个数量级以上，并将在各种动态成像问题中得到应用，特别是在对大脑功能活动的实时成像中。

As functional MRI provides a new way to study human brain functions, it becomes increasingly important to improve the temporal resolution of the experiments so that rapid transient neuronal events can be captured. In conventional Fourier transform-based imaging methods, each of the dynamic images is acquired independently so that the temporal resolution possible is limited by the number of data points collected in Fourier space to achieve a certain level of spatial resolution. This research seeks to overcome this problem by incorporating a priori information into the imaging process. Specifically, new optimal methods for data acquisition and image reconstruction are being developed that exploit both the underlying model for the signal as well as the available a priori information about both the object being imaged and the dynamic changes. These methods promise to give more that an order of magnitude of improvement in imaging efficiency over the traditional Fourier imaging methods, and should find application in a variety of dynamic imaging problems, particularly in real-time imaging of functional brain activities.