RATIONALE AND OBJECTIVES: Geometrical distortion is a well-known problem in structural magnetic resonance imaging (MRI), leading to pixel shifts with variations up to several millimeters. Because the main factors of geometrical distortion are proportional to B(0), MRI spatial encoding distortions tend to increase with higher magnetic field strength. With the increasing prospects of utilizing ultra-high-field MRI (B(0) ≥ 7 Tesla) for neuroimaging and subsequently for image-guided neurosurgical therapy, the evaluation and correction of geometrical distortions occurring in ultra-high-field MRI are essential preconditions for the integration of these data. Hence, we conducted a phantom study to determine hardware-related geometrical distortion in clinically relevant sequences for structural imaging at 7 T MRI and compared the findings to 1.5 T MRI. MATERIAL AND METHODS: Hardware-related geometrical distortion was evaluated using a MRI phantom (Elekta, Sweden). Both applied scanner systems (Magnetom Avanto 1.5 T and Magnetom 7 T, Siemens Healthcare, Erlangen, Germany) were equipped with similar gradient coils capable of delivering 45 mT/m of maximum amplitude and a slew rate of 220 mT/m/ms. Distortion analysis was performed for various clinically relevant gradient echo and spin echo sequences. RESULTS: Overall, we found very low mean geometrical distortions at both 7 T and 1.5 T, although single values of up to 1.6 mm were detected. No major differences in mean distortion between the sequences could be found, except significantly higher distortions in turbo spin-echo sequences at 7 T, mainly caused by B(1) inhomogeneities. CONCLUSION: Hardware-related geometrical distortions at 7 T MRI are relatively small, which may be acceptable for image coregistration or for direct tissue-targeting procedures. Using a subject-specific correction of object-related distortions, an integration of 7 T MRI data into image-guided applications may be feasible.
RATIONALE AND OBJECTIVES: Geometrical distortion is a well-known problem in structural magnetic resonance imaging (MRI), leading to pixel shifts with variations up to several millimeters. Because the main factors of geometrical distortion are proportional to B(0), MRI spatial encoding distortions tend to increase with higher magnetic field strength. With the increasing prospects of utilizing ultra-high-field MRI (B(0) ≥ 7 Tesla) for neuroimaging and subsequently for image-guided neurosurgical therapy, the evaluation and correction of geometrical distortions occurring in ultra-high-field MRI are essential preconditions for the integration of these data. Hence, we conducted a phantom study to determine hardware-related geometrical distortion in clinically relevant sequences for structural imaging at 7 T MRI and compared the findings to 1.5 T MRI. MATERIAL AND METHODS: Hardware-related geometrical distortion was evaluated using a MRI phantom (Elekta, Sweden). Both applied scanner systems (Magnetom Avanto 1.5 T and Magnetom 7 T, Siemens Healthcare, Erlangen, Germany) were equipped with similar gradient coils capable of delivering 45 mT/m of maximum amplitude and a slew rate of 220 mT/m/ms. Distortion analysis was performed for various clinically relevant gradient echo and spin echo sequences. RESULTS: Overall, we found very low mean geometrical distortions at both 7 T and 1.5 T, although single values of up to 1.6 mm were detected. No major differences in mean distortion between the sequences could be found, except significantly higher distortions in turbo spin-echo sequences at 7 T, mainly caused by B(1) inhomogeneities. CONCLUSION: Hardware-related geometrical distortions at 7 T MRI are relatively small, which may be acceptable for image coregistration or for direct tissue-targeting procedures. Using a subject-specific correction of object-related distortions, an integration of 7 T MRI data into image-guided applications may be feasible.
Authors: Inge Compter; Jurgen Peerlings; Daniëlle B P Eekers; Alida A Postma; Dimo Ivanov; Christopher J Wiggins; Pieter Kubben; Benno Küsters; Pieter Wesseling; Linda Ackermans; Olaf E M G Schijns; Philippe Lambin; Aswin L Hoffmann Journal: MAGMA Date: 2016-03-30 Impact factor: 2.310
Authors: Alexandre Boutet; Robert Gramer; Christopher J Steele; Gavin J B Elias; Jürgen Germann; Ricardo Maciel; Walter Kucharczyk; Ludvic Zrinzo; Andres M Lozano; Alfonso Fasano Journal: Curr Neurol Neurosci Rep Date: 2019-05-30 Impact factor: 5.081
Authors: Birgit R Plantinga; Yasin Temel; Alard Roebroeck; Kâmil Uludağ; Dimo Ivanov; Mark L Kuijf; Bart M Ter Haar Romenij Journal: Front Hum Neurosci Date: 2014-11-05 Impact factor: 3.169
Authors: Anneke Alkemade; Gilles de Hollander; Max C Keuken; Andreas Schäfer; Derek V M Ott; Johannes Schwarz; David Weise; Sonja A Kotz; Birte U Forstmann Journal: PLoS One Date: 2017-04-19 Impact factor: 3.240
Authors: Jurgen Peerlings; Inge Compter; Fiere Janssen; Christopher J Wiggins; Alida A Postma; Felix M Mottaghy; Philippe Lambin; Aswin L Hoffmann Journal: Phys Imaging Radiat Oncol Date: 2019-01-03