Maryna Waszak1,2,3, Pavel Falkovskiy1,2,3, Tom Hilbert1,2,3, Guillaume Bonnier1,2,3, Bénédicte Maréchal1,2,3, Reto Meuli3, Rolf Gruetter3,4,5, Tobias Kober1,2,3, Gunnar Krueger2,3,6. 1. Advanced Clinical Imaging Technology (HC CMEA SUI DI BM PI), Siemens Healthcare AG, Lausanne, Switzerland. 2. LTS5, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland. 3. Department of Radiology, University Hospital (CHUV), Lausanne, Switzerland. 4. Centre d'Imagerie BioMedicale (CIBM), École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland. 5. Department of Radiology, University of Geneva, Geneva, Switzerland. 6. Siemens Medical Solutions USA, Inc, Boston MA, USA.
Abstract
PURPOSE: We suggest a motion correction concept that employs free-induction-decay (FID) navigator signals to continuously monitor motion and to guide the acquisition of image navigators for prospective motion correction following motion detection. METHODS: Motion causes out-of-range signal changes in FID time series that, and in this approach, initiate the acquisition of an image navigator. Co-registration of the image navigator to a reference provides rigid-body-motion parameters to facilitate prospective motion correction. Both FID and image navigator are integrated into a prototype magnetization-prepared rapid gradient-echo (MPRAGE) sequence. The performance of the method is investigated using image quality metrics and the consistency of brain volume measurements. RESULTS: Ten healthy subjects were scanned (a) while performing head movements (nodding, shaking, and moving in z-direction) and (b) to assess the co-registration performance. Mean absolute errors of 0.27 ± 0.38 mm and 0.19 ± 0.24° for translation and rotation parameters were measured. Image quality was qualitatively improved after correction. Significant improvements were observed in automated image quality measures and for most quantitative brain volume computations after correction. CONCLUSION: The presented method provides high sensitivity to detect head motion while minimizing the time invested in acquiring navigator images. Limits of this implementation arise from temporal resolution to detect motion, false-positive alarms, and registration accuracy. Magn Reson Med 78:193-203, 2017.
PURPOSE: We suggest a motion correction concept that employs free-induction-decay (FID) navigator signals to continuously monitor motion and to guide the acquisition of image navigators for prospective motion correction following motion detection. METHODS: Motion causes out-of-range signal changes in FID time series that, and in this approach, initiate the acquisition of an image navigator. Co-registration of the image navigator to a reference provides rigid-body-motion parameters to facilitate prospective motion correction. Both FID and image navigator are integrated into a prototype magnetization-prepared rapid gradient-echo (MPRAGE) sequence. The performance of the method is investigated using image quality metrics and the consistency of brain volume measurements. RESULTS: Ten healthy subjects were scanned (a) while performing head movements (nodding, shaking, and moving in z-direction) and (b) to assess the co-registration performance. Mean absolute errors of 0.27 ± 0.38 mm and 0.19 ± 0.24° for translation and rotation parameters were measured. Image quality was qualitatively improved after correction. Significant improvements were observed in automated image quality measures and for most quantitative brain volume computations after correction. CONCLUSION: The presented method provides high sensitivity to detect head motion while minimizing the time invested in acquiring navigator images. Limits of this implementation arise from temporal resolution to detect motion, false-positive alarms, and registration accuracy. Magn Reson Med 78:193-203, 2017.