Deep learning tools for any anatomy

MR imaging is a dynamic field that is constantly evolving, with new applications and imaging markets emerging. Some of these applications come with substantial technical challenges, such as cardiac imaging, which has traditionally been challenged by phase artifacts and the need for rapid image acquisition. Deep-learning reconstruction methods, such as AIR™ Recon DL, have helped to increase spatial resolution, while allowing for shorter image acquisition times. Beyond AIR™ Recon DL, Sonic DL™ is GE HealthCare’s latest and most rapid MR acceleration technique that can achieve substantial reduction in scan time of cine acquisitions in cardiac imaging – up to 83% versus a fully sampled acquisition. In this issue of SIGNA™ Pulse of MR, Dr. Makoto Orii from Iwate Medical University in Japan relates his initial experience with Sonic DL™ Cine for evaluating ventricular function in the expanding market of cardiac MR. Dr. Melany Atkins from Fairfax Radiological Consultants in the US also discusses the impact of Sonic DL™ in cardiac cine acquisitions in evaluating patients with cardiac dysrhythmias.