View Presentation Video

View Presentation Video

Keywords: Image acquisition: Motion correction

A number of techniques have been proposed to address the motion problem in MRI. One such technique is the self-gating strategy, which attempts to take advantage of MR data itself to deal with motion. This lecture will cover the key concepts in self-gating strategy and how to extract self-gating signals with different sampling patterns. This lecture will especially focus on cardiac applications, where both respiratory and cardiac motion exist, and will also cover the state-of-the-art self-gating techniques when signal intensity is changing during acquisition.

View Presentation Video

Keywords: Image acquisition: Fast imaging

This tutorial will focus on discussing various dynamic imaging models for accelerated imaging. The models will be discussed in a unified perspective. Linear models such as view-sharing, UNFOLD, k-t BLAST,partial separability model (low-rank model) will first be discussed. Non-linear models such as compressed sensing, joint low rank and sparsity based, blind compressed sensing will then be reviewed. Imaging models without the need of explicit deformation estimation to perform motion resolved reconstruction such as extra-dimensional based models , manifold learning models will also be reviewed. Several application examples in free breathing cardiac, real time speech, free breathing liver DCE-MRI will be highlighted.

View Presentation Video

Keywords: Image acquisition: Modelling, Image acquisition: Motion correction, Image acquisition: Reconstruction

This educational lecture covers mathematical models used to describe patient motion, how different models are used for different applications, and connects motion modelling to data acquisition and reconstruction. The talk will begin with an overview of model based image reconstruction (MBIR) assuming no motion. Next, we will cover choices for modeling motion within MBIR, including rigid vs. non-rigid, types of k-space navigators, temporal modeling (intra-image vs. inter-image), and periodic vs. non-periodic motion. The talk will also discuss how each of these choices affect the objective functions, regularizers, and optimization algorithms used in the iterative model-based image reconstruction.  

View Presentation Video

Keywords: Image acquisition: Fast imaging, Image acquisition: Reconstruction, Image acquisition: Machine learning

MR signature matching (MRSIGMA) enables to perform real-time 4D MRI to guide radiotherapy using a combined MR-linac system. MRSIGMA shifts the acquisition and reconstruction burden to a motion learning step, where a 4D motion dictionary of 3D motion states and corresponding motion signatures is learned for each treatment fraction. Once the 4D motion dictionary is learned, fast signature-only acquisition and matching can be performed to minimize imaging latency and obtain 3D images in less than 300ms. The lecture will discuss acquisition, reconstruction and deep learning techniques to implement signature matching for radiotherapy monitoring, adaptation and dose calculation in in real-time.

View Presentation Video

Keywords: Image acquisition: Fast imaging, Image acquisition: Modelling, Image acquisition: Reconstruction

In this lecture, I will sketch the main traits of real-time motion estimation in MRI. After a brief overview of the actual and envisioned applications, I will review the main techniques involved in the acquisition, reconstruction and post-processing steps. These can be subdivided in two categories, namely: indirect methods, where motion is estimated upon registration of images, and direct methods, where motion is reconstructed at once from the k-space data. Recent machine-learning solutions will be reviewed as well.

View Presentation Video

Keywords: Image acquisition: Motion correction, Cardiovascular: Cardiac, Image acquisition: Reconstruction

This tutorial will focus on how to implement a framework for retrospective respiratory motion correction in practice. We will focus on whole-heart cardiac MRI and will review the steps required to produce motion compensated images, including key components in the image acquisition sequence to enable measuring respiratory motion directly, alternatives for estimating rigid and non-rigid motion, and ways of including motion information in the image reconstruction process. The talk will demonstrate some challenging cases and the impact of choosing rigid/non-rigid respiratory motion models for a variety of whole-heart applications.