Fast Regularized Parallel Imaging in an MR Image-Guided Therapy Application

We present an overview of parallel MR imaging and its use in an image-guided therapy application. Different parallel imaging methods have different strengths, which we combine to increase MR image acquisition efficiency. As imaging efficiency increases, the image reconstruction problem often becomes poorly conditioned. The conditioning of reconstruction problems can be improved through the use of solution-space regularization. We review our fast iterative regularized reconstruction method, our implementation of this method on a multi-node computer, and demonstrate its use in high efficiency MR imaging for a cardiac therapy application.