First Real-Time MRI of Moving Organs and Joints, Max Planck Institute Study Published in NMR in Biomedicine
Published: Aug 31, 2010
MedGadget -- Moving MRI images of the heart and other organs are nothing new. However, up until now these were always reconstructed afterwards by combining signals and images from the same cardiac phase acquired during breath-holding into the resulting images. Researchers at the Max Planck in Göttingen have now proven it is possible to acquire images in real-time with up to fifty frames per second. Previously it was only possible to acquire images at a maximum rate of about one per second. They have used a combination of fast low angle shot (FLASH) gradient-echo imaging sequences, radial encoding and an iterative image reconstruction by regularized nonlinear inversion to greatly speed up imaging acquisition. The main improvement lays in the iterative reconstruction which has previously been used for reducing CT radiation dose, but has now shown to be very useful in MRI as well. For example, the images of a cross-section of the heart shown above were acquired at 33 milliseconds per image giving a framerate of 30 images per second, with an in-plane resolution of 1.5 millimeters and a slice thickness of 8 millimeters. Linked in the press release is a video of the beating heart and an impressive movie of movements during speech production. The possibility to do real-time dynamic imaging opens up a lot of new opportunities for MR imaging, including, for example, swallowing studies, cardiac and abdominal imaging without breath-holding and interventional procedures using MRI. In principle this technique is possible to implement on most current scanners, however the main limitation is the amount of computing power required to perform the real-time reconstruction. Currently one minute of images requires half an hour to process. However, the authors expect improvements in the image reconstruction algorithms for parallelized GPU's and other improvements to reduce reconstruction time and improve image quality. The results are published online in the journal NMR in Biomedicine.