Evaluating the role of 7-Tesla magnetic resonance imaging in neurosurgery: Trends in literature since clinical approval

Table 1 Summary of studies included in the review falling under each neurosurgically treated pathology

Pathology	Ref.	Aim of the study	Applied 7T MRI protocol
Epilepsy	Stefanits et al[10]	To correlate noninvasive, high-resolution, morphological 7T MRI of the hippocampus in temporal lobe epilepsy (TLE) patients with histopathological findings	T2-weighted (T2w) 2D fast spin echo (FSE) sequence, obtained in paracoronal, hippocampal plane perpendicular to the central sulcus (matrix, 688 × 688; field of view (FOV), 230 × 172.5; image resolution, 0.33 mm × 0.33 mm × 1.5 mm; slices, 25; parallel imaging, 2; repetition time (TR), 4500 milliseconds (ms); echo time (TE), 81 ms) with an acquisition time of 8 min 48 s
	Wang et al[12]	To assess the clinical value of in vivo structural 7T MRI and its post-processing in patients with pharmacoresistant epilepsy who underwent presurgical evaluation and had a nonlesional 3T MRI scan	A standard epilepsy protocol was used with the following sequences on a 7T MRI scanner (Magnetom, Siemens, Erlangen, Germany) with a head-only circularly polarized transmit and 32-channel phased array receive coil (Nova Medical, Wilminutesgton, MA): 3D T1-MP2RAGE: Sagittal acquisition, TR/TE = 6000/3 ms, TI1/TI2 = 700/2700 ms, flip angle (FA) 1/FA 2 = 4°/5°, 0.75 mm isotropic-voxel resolution, 208 slices, total acquisition time (TA) = 9 min 32 s; 2D T2-GRE (spoiled-gradient echo): Axial and oblique coronal acquisition, TR/TE = 2290/17.8 ms, FA = 23°, in-plane resolution = 0.38 mm × 0.38 mm, slice thickness = 1.5 mm, 60 slices, no gap, TA = 9 min 50 s; 2D FLAIR: Axial and oblique coronal acquisition, TR/TE = 9000/124 ms, TI = 2600 ms, in-plane resolution = 0.75 mm × 0.75 mm, slice thickness = 2 mm, 45 slices, 30% gap, TA = 3 min 2 s; 3D susceptibility weighted imaging (SWI) (included only for selected cases such as vascular malformation): TR/TE = 23/15 ms, FA = 20°, voxel size = 0.49 mm × 0.49 mm × 0.8 mm, 144 slices, TA = 8 min 16 s. Two dielectric calcium titanate pads with passive B1 shimming were used to improve the signal loss in the temporal lobes
	Rutland et al[15]	To perform hippocampal subfield-specific tractography and quantify connectivity of the subfields in MRI-negative patients. Abnormal connectivity of the hippocampal subfields may help inform seizure focus hypothesis and provide information to guide surgical intervention	Participants were scanned under an Institutional Review Board-approved protocol using a 7T whole body scanner. A SC72CD gradient coil was used (Gmax = 70 mT/m, max slew rate = 200 T/m/s), with a single channel transmit and 32 channel receive head coil. The MRI scan included a T1-weighted MP2RAGE sequence: TR = 6000 ms, TE = 3.62 ms, FA = 5°, FOV = 240 mm × 320 mm, slices = 240, 0.7 mm3 isotropic resolution, scan time = 7 min 26 s. A coronal-oblique T2w turbo spin echo sequence was included: TR = 6900 ms, TE = 69 ms, FA = 150°, FOV = 202 mm × 202 mm, in-plane resolution 0.4 mm × 0.4 mm, slice thickness = 2 mm, slices = 40, time = 6 min 14 s. A high-angular-resolved diffusion-weighted imaging (HARDI) dMRI sequence was also performed with whole-brain coverage: b = 1200 s/mm2, TR = 7200 ms, TE = 67.6 ms, 1.05 mm isotropic resolution, in-plane acceleration R = 3 (GRAPPA), reversed phase encoding in the AP and PA direction for paired acquisition in 68 directions, with a TA of 20 min
	Veersema et al[13]	To determine whether the use of 7T MRI in clinical practice leads to higher detection rates of focal cortical dysplasias in possible candidates for epilepsy surgery	7T MRI parameters are as follows: 3D T1w TFE: TE = 2.9 ms, TI = 1200 ms, TR = 9 ms, FA = 6°, resolution = 0.81 mm × 0.81 mm × 0.80 mm, matrix = 248 × 312 × 475, acquisition time = 9 min 34 s. 3D T2w turbo spin echo (TSE): TE = 302 ms, TR = 3200 ms, FA = 90°, resolution = 0.70 mm × 0.70 mm × 0.70 mm, matrix = 356 × 357 × 543, acquisition time = 10 min 24 s. 3D T2*w: TE = 27 ms, TR = 88 ms, FA = 24°, resolution = 0.50 mm × 0.50 mm × 0.50 mm, matrix = 480 × 381 × 600, acquisition time = 7 min 23 s; 3D MP-FLAIR: TE = 300 ms, TI = 2200 ms, TR = 8000 ms, FA = 90°, resolution = 0.80 mm × 0.82 mm × 1.00 mm, matrix = 312 × 304 × 380, acquisition time = 8 min 40 s. 3D WMS: TE = 2.0 ms, TI = 600 ms, TR = 6.73 ms, FA = 4.5°, resolution = 0.80 mm × 0.80 mm × 0.80 mm, matrix = 320 × 320 × 474, acquisition time = 8 min 42 s
	Zhang et al[11]	To compare the hippocampal internal architecture (HIA) between 3 and 7T MRI in patients with TLE	All patients underwent MRI scans with a 7 T scanner. Thirty-two channel head coils were used with both scanners. For the assessment of HIA, T2w images (T2WI) in the coronal plane located perpendicular to the long axis of the hippocampus were collected (7T T2WI-TSE): TR = 9640 ms, TE = 72 ms, resolution 0.3 mm × 0.3 mm × 2.0 mm, FA = 60°, TA = 11 min 26 s. 7T T1WI 3D-magnetization prepared rapid acquisition gradient echo (3D-MPRAGE): TR = 2200 ms, TE = 2.98 ms, resolution = 0.7 mm × 0.7 mm × 0.7 mm, FA = 8°, TA = 10 mi 16 s
	Sharma et al[14]	To quantitatively assess surgical outcomes in epilepsy patients who underwent scanning at 7T whose lesions were undetectable at conventional field strengths (1.5T/3T)	Unspecified
	van Lanen et al[16]	To assess whether 7T MRI increases the sensitivity to detect epileptogenic lesions	N/A (systematic review)
Pituitary adenoma	Yao et al[18]	To examine the utility of 7T MRI in predicting the tumor consistency of pituitary adenomas	High-resolution 7T TSE (0.4 mm × 0.4 mm × 2 mm), MP2-RAGE (0.75 mm isotropic), and TOF (0.26 mm × 0.26 mm × 0.4 mm) acquisitions on all patients
	Rutland et al[20]	To investigate microstructural damage caused by pituitary macroadenomas by performing probabilistic tractography of the optic tracts and radiations using 7T diffusion-weighted MRI (DWI). These imaging findings were correlated with neuro-ophthalmological results to assess the utility of ultra-high-field MRI for objective evaluation of damage to the anterior and posterior visual pathways	Participants were scanned under an Institutional Review Board-approved protocol using a 7T whole body scanner. A SC72CD gradient coil was used (Gmax = 70 mT/m, max slew rate = 200 T/m/s), with a single channel transmit and 32-channel receive head coil. Scanning included a T1-weighted MP2RAGE sequence with the following parameters: TE (ms) = 5.1, TR (ms) = 6000, TI (ms) = 1050 (3000), FA = 5° (4°), FOV = 240 mm × 320 mm, slices = 240, resolution = 0.7 mm isotropic, scan time = 7 min 26 s. Quantitative T1-maps were derived from the MP2RAGE sequence. A coronal-oblique T2-TSE (TE = 69, TR = 6900, FA = 150°, FOV = 202 mm × 202 mm, in-plane resolution 0.4 mm × 0.4 mm, slice thickness = 2 mm, slices = 40, time = 6 min 14 s) and HARDI dMRI (b = 1200 mm2/sec, TE = 67.6, TR = 7200, resolution = 1.05 mm isotropic, in-plane acceleration R = 3 (GRAPPA), reversed phase encoding in AP and PA directions for paired acquisition in 68 directions, TA = 20) sequences were acquired
	Rutland et al[21]	To leverage ultra-high-field 7T MRI to study the retinotopic organization of the primary visual cortex (V1) and correlate visual defects with cortical thinning in V1 to characterize consequences of pituitary adenomas on the posterior visual system	Participants were scanned using a 7T whole-body scanner (Magnetom, Siemens Healthineers). A SC72CD gradient coil was used (Gmax = 70 mT/m, maximum slew rate = 200 T/m/sec), with a single-channel transmit and 32-channel receive head coil (Nova Medical). Scanning included a T1-weighted MP2RAGE sequence with the following parameters: TE = 3.62 ms, TR 6000 ms, TI = 1050 ms (2nd pulse 3000 ms), FA = 5° (2nd pulse 4°), FOV = 224 mm × 168 mm, number of slices = 240, voxel size = 0.7 mm isotropic, and scan time = 8 min 8 s. Coronal oblique T2w turbo spin echo (TE = 59 ms, TR = 6000 ms, FA = 180°, FOV = 200 mm × 168 mm, in-plane voxel size = 0.4 mm × 0.4 mm, slice thickness = 2 mm, number of slices = 60, time = 6 min 50 s) and high-angular-resolved DWI (b = 1200 mm2/s, TE = 67.6 ms, TR = 7200 ms, voxel size = 1.05 mm isotropic, FA = 180°, number of slices = 66, in-plane acceleration R = 3, reversed-phase encoding in anteroposterior and posteroanterior directions for paired acquisition in 64 directions, TA = 18 min 38 s) sequences were prescribed. Dielectric pads and localized shimming methods were employed to reduce signal artifact at the skull base
	Patel et al[22]	To describe the initial experience using ultra-high-field 7T MRI in patients with suspected Cushing’s disease and negative or equivocal imaging at conventional field strengths	Patients were scanned on a Siemens Terra 7T system using a Nova Medical 1Tx/32Rx head coil. Pre- and post- contrast (0.2 mL/kg gadoterate megluminutese) T1-weighted pituitary sequences included coronal and sagittal 2D TSE (TR = 960 ms, TE = 10 ms, voxel size = 0.2 mm × 0.2 mm × 2.0 mm), 3D SPACE (TR = 1200 ms, TE = 12 ms, variable FA, voxel size = 0.5 mm × 0.5 mm × 0.5 mm), and 3D MPRAGE (TR = 2300 ms, TE = 2.95 ms, FA = 7°, voxel size = 0.7 mm × 0.7 mm × 0.7 mm). Not all patients were scanned with all sequences due to changes in clinical protocol during the study period
	Rutland et al[23]	To examine 7T DWI as a novel method of measuring the consistency of pituitary adenomas	Participants were scanned using a 7T whole-body scanner. A SC72CD gradient coil was used (Gmax = 70 mT/m, maximum slew rate = 200 T/m/s), with a single-channel transmit and 32-channel receive head coil. Scanning included a T1-weighted MP2RAGE sequence with the following parameters: TE = 3.62 ms, TR = 6000 ms, TI = 1050 ms (2nd pulse 3000 ms), FA = 5° (2nd pulse 4°), FOV = 224 mm × 168 mm, number of slices = 240, voxel size = 0.7 mm isotropic, and scan time = 8 min 8 s. Coronal oblique T2w turbo spin echo (TE = 59 ms, TR = 6000 ms, FA = 180°, FOV = 200 mm × 168 mm, in-plane voxel size = 0.4 mm × 0.4 mm, slice thickness = 2 mm, number of slices = 60, time = 6 min 50 s) and high-angular-resolved DWI (b = 1200 mm2/s, TE = 67.6 ms, TR = 7200 ms, voxel size = 1.05 mm isotropic, FA = 180°, number of slices = 66, in-plane acceleration R = 3, reversed-phase encoding in anteroposterior and posteroanterior directions for paired acquisition in 64 directions, TA = 18 min 38 s) sequences were prescribed. Dielectric pads and localized shimming methods were employed to reduce signal artifact at the skull base
	Rutland et al[24]	To determine the efficacy of 7T MRI in identifying radiological markers for endocrine function	The Institutional Review Board-approved protocol employed a 7T whole-body MRI scanner (Magnetom, Siemens Healthcare, Erlangen, Germany) with a SC72CD gradient coil (Gmax = 70 mT/m, max slew rate = 200 T/m/s), and used a single channel transmit and 32-channel receive head coil (Nova Medical, Wilminutesgton, MA, United States). Sequences included a T1-weighted MP2RAGE[14] sequence: TE (ms) = 3.62, TR (ms) = 6000, TI (ms) = 1050/3000, FA = 5°/4°, FOV = 224 mm × 168 mm, slices = 240, resolution = 0.7 mm isotropic, scan time = 8 min 8 s. Quantitative T1-maps were derived from the MP2RAGE sequence. Coronal-oblique and axial T2-TSE (TE = 60 ms, TR = 6000 ms, FA = 180°, FOV = 200 mm × 168 mm, in-plane resolution 0.4 mm × 0.4 mm, slice thickness = 2 mm, slices = 60, time = 6 min 50 s) sequences were acquired. A T1-weighted MPRAGE sequence was also obtained (TE = 4.1 ms, TR = 3000 ms, TI = 1050 ms, FA = 7°, resolution = 0.7 mm isotropic, scan time = 7 min 40 s)
	Rutland et al[17]	The study quantifies visualization of tumor features and adjacent skull base anatomy in a homogeneous cohort of pituitary adenoma patients comparing 7T MRI vs standard lower field MRI	Unspecified
Parkinson`s disease	Shamir et al[30]	To validate the clinical application accuracy of the 7T-ML method by comparing it with identification of the subthalamic nucleus (STN) based on intraoperative microelectrode recordings	Unspecified
	Lau et al[27]	To integrate ultra-high-field template data into the clinical workflow to assist with target selection in deep brain stimulation (DBS) surgical planning	Patients were scanned on a 7T scanner (Agilent, Santa Clara, California, United States/Siemens, Erlangen, Germany) via a 24-channel transmit-receive head coil array constructed in-house with a receiver bandwidth of 50 kHz. A T1w MPRAGE sequence was acquired (TR = 8.1 ms, TE = 2.8 ms, inversion time (TI) = 650 ms, FA = 11°, matrix: 256 × 512, 230 slices, resolution = 0.59 mm × 0.43 mm × 0.75 mm). Then, a T2w turbo spin-echo 3D (TR = 3D sagittal, matrix: 260 × 366, 266 slices, resolution = 0.6 mm3, 4 averages) was acquired. High- resolution in vivo templates were created by performing group-wise linear and nonlinear registration of 12 normal subjects scanned on a human 7T imager with both T1w and T2w contrasts (available for download at http://www.nitrc.org/projects/deepbrain7t/) resulting in an unbiased group nonlinear T1w average and T2w averages at submillimeter resolution
	La et al[26]	To measure hippocampal subfields in vivo using ultra-high-field 7T MRI and determine if these measures predict episodic memory impairment in PD during life	Participants were scanned with a 7T GE Healthcare Discovery MR950 MRI whole-body scanner (GE Healthcare, Waukesha, WI) using a 32-channel radiofrequency (RF) receive head coil contained within a quadrature transmit coil (Nova Medical, Inc., Wilminutesgton, MA). Sixteen oblique coronal images oriented perpendicular to the longitudinal axis of the hippocampus were acquired with a T2w FSE sequence: TE = 47 ms; TR = 5-8 s (cardiac gated); acquired voxel size was 0.22 mm × 0.22 mm × 1.5 mm with a slice gap of 0.5 mm, interpolated by zero filling to 0.166 mm × 0.166 mm × 1.5 mm
	Lau et al[27]	To integrate ultra-high-field template data into the clinical workflow to assist with target selection in DBS surgical planning	Patients were scanned on a 7T imager (Agilent, Santa Clara, California, United States/Siemens, Erlangen, Germany) via a 24-channel transmit-receive head coil array constructed in-house with a receiver bandwidth of 50 kHz. A T1w MPRAGE sequence was acquired (TR = 8.1 ms, TE = 2.8 ms, TI = 650 ms, FA = 11°, matrix = 256 × 512, 230 slices, resolution = 0.59 mm × 0.43 mm × 0.75 mm). Then, a T2w turbo spin-echo 3D (TR 3D sagittal, matrix: 260 × 366, 266 slices, resolution = 0.6 mm3, 4 averages) was acquired. High-resolution in vivo templates were created by performing group-wise linear and nonlinear registration of 12 normal subjects scanned on a human 7T imager with both T1w and T2w contrasts (available for download at http://www.nitrc.org/projects/deepbrain7t/) resulting in an unbiased group nonlinear T1w average and T2w averages at submillimeter resolution
	La et al[26]	To measure hippocampal subfields in vivo using ultra-high field 7T MRI and determine if these measures predict episodic memory impairment in PD during life	Participants were scanned with a 7T GE Healthcare Discovery MR950 MRI whole-body scanner (GE Healthcare, Waukesha, WI) using a 32-channel RF receive head coil contained within a quadrature transmit coil (Nova Medical, Inc., Wilminutesgton, MA). Sixteen oblique coronal images oriented perpendicular to the longitudinal axis of the hippocampus were acquired with a T2w FSE sequence: TE = 47 ms; TR = 5-8 s (cardiac gated); acquired voxel size was 0.22 mm × 0.22 mm × 1.5 mm with a slice gap of 0.5 mm, interpolated by zero filling to 0.166 mm × 0.166 mm × 1.5 mm
	Oh et al[25]	To investigate patterns in gray matter changes in patients with Parkinson's disease by using an automated segmentation method with 7T MRI	High-resolution T1-weighted 7T MRI volumes of 24 hemispheres were acquired from 12 Parkinson's disease patients. Magnetic resonance images from all subjects were acquired using a 7T MRI system (Philips Healthcare, Cleveland, OH, United States) with a 32-channel head coil (Nova Medical, Wilminutesgton, MA, United States). Three-dimensional anatomical brain scans were acquired using a magnetization-prepared rapid-acquisition gradient-echo sequence-induced T1-weighted imaging system with the following settings: TR = 4.4 ms, TE = 2.2 ms, slice thickness = 0.5 mm, in-plane resolution = 0.5 mm × 0.5 mm, matrix size = 432 × 432, number of axial slices = 432, and TA = 5 min 57 s
	Mathiopoulou et al[29]	To explore whether combining 7T T2w and DWI sequences allows for selective segmenting of the motor part of the STN and, thus, for possible optimization of DBS	7T T2w and DWI sequences were obtained, and probabilistic segmentation of motor, associative, and limbic STN segments was performed. The MR-sequences were acquired before surgery on a 3T Philips Ingenia scanner (Philips Healthcare, Best, The Netherlands): (1) 3D sagittal T1-weighted, gadolinium-enhanced (TR = 8.81 ms, TE = 4.03 ms, echo train length (ETL) = 242, FOV = 256 mm, slice thickness = 0.9 mm, scan time = 8 min); (2) 3D axial T2w (TR = 2500 ms, TE = 230 ms, ETL = 133, FOV = 250 mm, slice thickness = 1.1 mm, scan time = 3 min); and (3) DWI (TR = 8234 ms, TE = 96 ms, b = 1200 s/mm2, 32 gradient directions, phase encoding anterior-posterior, no reverse encoding, FOV = 256 mm, slice thickness = 2.0 mm, scan time = 14 min). The following MR sequences were acquired on a 7T Achieva system (Philips Healthcare) using a 32 channel receive coil (Nova Medical, Wilminutesgton, MA): (1) 3D sagittal T2w with Turbo spin echo imaging (TR = 3000 ms, TE = 324 ms, ETL = 182, FOV = 250 × 250 × 190, FA = 100°, voxel size = 0.7 mm isotropic, scan duration = 7 min); and (2) DWI (TR = 6084 ms, TE = 70 ms, ETL = 59, b = 1000 s/mm2, 32 directions, FOV = 140 × 177 × 110, voxel size = 1.4 mm × 1.4 mm in-plane, slice thickness = 1.5 mm, scan duration = 13 min)
	Isaacs et al[28]	The study aimed to test whether optimized 7T imaging protocols result in less variable targeting of the STN for DBS compared to clinically utilized 3T images	The 7T scan was acquired with a Siemens Magnetom scanner using a 32-channel head coil at the Scannexus Centre for Neuroimaging in Maastricht. Whole brain T1w 3D images were obtained with an adapted version of the multi echo MP2RAGE (magnetization-prepared rapid gradient echo multi-echo) sequence with 0.8 mm isotropic voxel sizes and the following parameters: 208 slices, TR = 6000 ms, TE 1/2 =(2.74 ms/8.71 ms), TI 1/2 = (750 ms/29000 ms), FA 1/2 =(4°/6°), BW 1/2 =(350 Hz/Px/150 Hz/Px), ES =13.6 ms, interleaved and single shot multi slice mode and interleaved, sagittal orientation acquisition in the anterior-posterior direction, phase partial Fourier 6/8, parallel acquisition with GRAPPA and acceleration factor of 3 and TA of 10 min 56 s. Where possible, dielectric pads were placed between the side of the participants head and the receiver coil to reduce B1 inhomogeneity artefacts. The T2w 3D scan was acquired with a partial volume gradient echo ASPIRE (multi-channel phase data from multi-echo acquisitions) sequence covering the subcortex with 0.5 mm isotropic voxel sizes and the following parameters: 90 slices, 16.7% slice oversampling, TR = 33 ms, TE 1-4 =(2.49 ms, 6.75 ms, 13.50 ms, 20.75 ms), FA =12°, BW 1-4 = (300 Hz/px, 300 Hz/px, 200 Hz/px, 100 Hz/px), interleaved multi slice mode, sagittal orientation acquisition in the anterior-posterior direction, slice partial Fourier 7/8, parallel acquisition with GRAPPA and acceleration factor of 2 and TA 7 min 42 s
Cerebrovascular diseases	Van Tuijl et al[34]	To gain insight into hemodynamic imaging markers of the CoW for unruptured intracranial aneurysm (UIA) development by comparing these outcomes to the corresponding contralateral artery without an UIA using 4D flow MRI	Participants underwent 7T MRI (Philips Healthcare, Best, The Netherlands) using a volume transmit and 32-channel receive coil (Nova Medical, Houston, United States). For this study, the Amsterdam Medical Center “PROspective Undersampling in multiple Dimensions” software patch was used, which enables a pseudospiral ky/kz-plane acquisition scheme designed for incoherent undersampling with a variable sampling density. Time-resolved 3D phase-contrast velocity maps (4D flow scan) over the cardiac cycle were acquired with the following parameters: Angulated coronal FOV 250 (feet–head) × 190 (right–left) × 20 (anterior-posterior) mm3, acquired resolution of 0.7 mm × 0.7 mm × 0.7 mm, repetition time/TE = 6.4/2.2, velocity encoding sensitivity 100 cm/s, FA = 15°, nominal acceleration factor 7, and 12 reconstructed cardiac phases (i.e., reconstructed temporal resolution 83 ms for a heart rate of 60 bpm). Retrospective gating used a peripheral pulse unit for heartbeat detection. Acquisition duration was 10 min
	Wrede et al[32]	To prospectively evaluate non-contrast-enhanced 7T magnetic resonance angiography (MRA) for delineation of UIAs	Both TOF MRA (time-of-flight magnetic resonance angiography) and non-contrast-enhanced MPRAGE were used at 7 T. 32-channel Tx/Rx RF head coil, maximum amplitude of 45 mT/m, slew rate of 200 mT/m/ms. Prior to acquisition of diagnostic sequences, B0-shimming was performed using a vendor provided gradient echo sequence and algorithm. For B1-field mapping and local FA optimization, a vendor-provided spin-echo type sequence was used; after a slice selective excitation, two refocusing pulses generate a spin-echo and a stimulated echo
	Uwano et al[36]	To investigate whether OEF maps generated by magnetic resonance quantitative susceptibility mapping (QSM) at 7 T enabled detection of OEF changes when compared with those obtained with PET	A 7T MRI scanner with quadrature transmission and 32-channel receive head coils was used. Source data of QSM were obtained using a 3-dimensional spoiled gradient recalled acquisition technique with the following scanning parameters: TR, 30 ms; TE, 15 ms; FA, 20°; FOV, 256 mm; acquisition matrix size, 512 × 256; slice thickness, 2 mm; number of slices, 160; reconstruction voxel size after zero-fill interpolation, 0.5 mm3; and scan time, 3 min 25 s. The sections were set in the orthogonal axial plane from the level of the superior cerebellar peduncle to high convexity. Magnitude as well as real/imaginary phase images were regenerated from this acquisition. Structural images including T2WI and magnetic resonance angiography were also obtained
	Uwano et al[37]	To examine whether whole-brain MRA at 7 T could non-invasively detect impaired CVR in patients with chronic cerebral ischemia by demonstrating the leptomeningeal collaterals	MRI examinations were performed with a 7T MRI scanner with a quadrature transmission and 32-channel receive head coil system. Whole-brain single-slab 3D time-of-flight MRA was performed using the following scanning parameters: TR, 12 ms; TE, 2.8 ms; FA, 12°; FOV, 22 cm; matrix size, 512 × 384; slice thickness, 0.5 mm (zero-fill interpolation); partition, 332; tilted optimized non-saturated excitation; number of excitations, 1; and acquisition time, 10 min 32 s. Subsequently, axial and coronal maximum intensity projection images were generated after skull stripping with SPM8 software. SPECT studies were performed using two scanners and iodine 123 N-isopropyl-p-iodoamphetaminutese (123I-IMP) at a resting state with the ACZ challenge, as described previously. The voxel size was 2.0 mm× 2.0 mm × 5.0 mm in scanner 1 and 1.72 mm × 1.72 mm × 1.72 mm in scanner 2. The cerebral blood flow was quantified using the 123I-IMP autoradiography method. Angiography and SPECT studies were performed within 10 d before/after the MRI examination
	Noureddine et al[35]	To evaluate RF-induced tissue heating around aneurysm clips during a 7T head MR examination and determine the decoupling distance between multiple implanted clips	Unspecified
	Millesi et al[33]	To delineate the wall of intracranial aneurysms to identify weak areas prone to rupture	Patients underwent ultra–high-field MRI that was performed in a whole-body 7T MRI unit (Siemens Healthcare) with a gradient strength of 40 mT/m using a 32-channel transmit/receive coil. The MRI sequence performed for this study was a nonenhanced MPRAGE sequence (TR/TE 3850/3.84 ms, isotropic voxel size 0.5 mm)
	Koemans et al[38]	To investigate whether a striped occipital cortex and intragyral hemorrhage, two markers recently detected on ultra-high-field 7T MRI in hereditary cerebral amyloid angiopathy (CAA), also occur in sporadic CAA (sCAA) or non-sCAA intracerebral hemorrhage (ICH)	T2-weighted images were obtained. MRI markers associated with small vessel disease were scored according to the Standards for Reporting Vascular changes on neuroimaging criteria
	Dammann et al[40]	To compare cerebral cavernous malformations -associated cerebral venous angioarchitecture between sporadic and familial cases using 7T MRI	In all patients, SWI was performed at 7 T (Magnetom 7T; Siemens Healthcare). Susceptibility-weighted imaging sequences were established in previous work. The whole-body ultra-high-field MR system was equipped with a single-channel transmitter/32-channel receiver head coil (Nova Medical), and a gradient system capable of 45 mT/m maximum amplitude and a slew rate of 220 mT/m/ms. The SWI parameters were as follows: TE 15 ms, TR 27 ms, FA 14°, in-plane resolution (R) 250 mm × 250 mm, slice thickness 1.5 mm, and bandwidth 140 Hz/pixel. The SWI data were processed to phase, magnitude, susceptibility, and minimum intensity projection images. In addition, anatomical T1w (TR 2500 ms, TE 1.4 ms, FA 6°, 0.7 mm isotropic), T2w (TR 6000 ms, TE 99 ms, FA 29°, ST 3 mm, R 0.5 mm2), and time-of-flight angiography (TR 20 ms, TE 4.3 ms, FA 20°, ST 0.4 mm, R 0.2 mm2) sequences were acquired
	Harteveld et al[31]	To compare 3T and 7T MRI in visualizing both the intracranial arterial vessel wall and vessel wall lesions	For 7T MRI, a whole-body system (Philips Healthcare, Cleveland, OH, United States) was used with a 32-channel receive coil and volume transmit/receive coil for transmission (Nova Medical, Wilminutesgton, MA, United States). The imaging protocol included a 3D whole-brain T1-weighted magnetisation-prepared inversion recovery turbo spin echo intracranial vessel wall sequence
	Jolink et al[39]	To assess presence and extent of contrast agent leakage distant from the hematoma as a marker of BBB disruption in patients with spontaneous ICH	7T MRI (Philips, Best, The Netherlands) scans were acquired with a standardized protocol; 3D T2w TR/equivalent TE = 3158/60 ms; voxel size acquired: 0.70 mm × 0.70 mm × 0.70 mm , reconstructed: 0.35 × mm × 0.35 mm × 0.35 mm), 3D T1-weighted (TR/TE 4.8/2.2 ms; voxel size acquired: 1.00 mm × 1.01 mm × 1.00 mm, reconstructed: 0.66 mm × 0.66 mm × 0.50 mm), dual echo 3D T2w (TR/first TE/ second TE 20/6.9/15.8 ms; voxel size acquired: 0.50 mm × 0.50 mm × 0.70 mm, reconstructed: 0.39 mm × 0.39 mm × 0.35 mm), and 3D FLAIR images were acquired (TR/TE/TI 8000/300/2325 ms; voxel size acquired: 0.80 mm × 0.82 mm × 0.80 mm , reconstructed: 0.49 mm × 0.49 mm × 0.40 mm). A gadolinium-containing contrast agent was administered in a single intravenous injection of 0.1 mL Gadovist/kg body weight with a maximum of 10 mL Gadovist or 0.2 mL Dotarem/kg body weight with a maximum of 30 mL Dotarem. Postgadolinium FLAIR images were acquired at least 10 min after contrast injection
TN	Moon et al[42]	To determine the central causal mechanisms of TN and the surrounding brain structure in healthy controls (HCs) and patients with TN using 7T MRI	Subjects underwent MRI scans using a 7T MR system (Philips Healthcare, Cleveland, OH, United States) with a 32-channel phased-array head coil (Nova Medical, Wilminutesgton, MA, United States). 3D anatomical brain scans were acquired using magnetization prepared rapid acquisition gradient echo (MP-RAGE) sequence-induced T1w imaging with the following settings: TR = 4.6 ms, TE = 2.3 ms, FA = 110°, slice thickness = 0.5 mm, in-plane resolution = 0.5 mm × 0.5 mm, matrix size = 488 × 396, number of axial slices = 320, and TA = 5 min 53 s
	Moon et al[41]	To investigate DTI parameters and the feasibility of DTI criteria for diagnosing TN	Imaging scans were acquired with a whole-body 7T MR system (Philips Healthcare, Cleveland, OH, United States) with a 16-channel receive head coil (Nova Medical, Wilminutesgton, MA, United States) and volume transmit. Standard scanning protocols included T1, 3D T2-VISTA, and DTI were used to assess microstructure and pathology changes. MRI sequences were: 7T T1 anatomical images (TR = 4.6 ms, TE = 2.3 ms, FA = 110°, thickness = 0.5 mm, voxel size 0.5 × 0.5, matrix size 488 × 396); three-dimensional T2-VISTA images to confirm offending vessels (TR = 2031 ms, TE = 303 ms, FA = 90°, voxel size 0.5 × 0.5, matrix size 360 × 360); and DTI (TR = 5606 ms, TE = 63 ms, FA = 90°, thickness = 1.5 mm, voxel size 1.5 × 1.8, matrix size 140 × 107, and b = 700 s/mm2)
Traumatic head injury	Hütter et al[43]	To evaluate the possible prognostic benefits of 7T SWI of traumatic cerebral microbleeds over 3T SWI to predict the acute clinical state and subjective impairments, including health-related quality of life, after closed head injury	The MRI examinations and their methods have been described in detail in an earlier publication. The 7T SWI acquisition time was about 13 min. UHF MR examinations were performed on a 7 T whole-body research system (Magnetom 7T, Siemens Healthcare, Germany) used in combination with 32-channel radiofrequency head coils
MS	Chou et al[44]	To evaluate the sensitivity of 7T magnetization-transfer-weighted (MTw) images in the detection of white matter lesions compared with 3T FLAIR	Participants underwent MRI in the Sir Peter Mansfield Imaging Centre using a 7T Philips Achieva scanner (Philips Medical Systems, Best, the Netherlands). The scanning protocols included 3D T1-weighted MP-RAGE imaging to assist with coregistering 3T 2D-FLAIR and 7T 3D-MTw images. Acquisition parameters at 7 T were TE = 3.2 ms; TR = 6.9 ms; TI = 800 ms; FA of the TFE readout pulse = 80°; TFE factor = 240; shot-to-shot interval = 8 s; spatial resolution = 1.25 mm × 1.25 mm × 1.25 mm; FOV = 200 mm × 200 mm × 72.5 mm; reconstruction matrix = 160 × 160 × 58 mm3; and TA = 2 min
	Choksi et al[45]	The study aimed to identify white matter tracts associated with the severity of neurogenic lower urinary tract dysfunction in women with MS	7T DTI images were acquired (matrix 158 × 158, slick thickness 1.4 mm, FOV 220 cm × 220 c2, 64 directions, b = 1000 s/mm2, total scan time 12 min and 14 s) on a 7T Siemens MAGNETOM Terra MRI scanner with a 32-channel single transmit head coil (3T Siemens MAGNETOM Vida MRI scanner was used in two subjects with contraindications for 7T scanner)
	Beck et al[46]	The study aimed to characterize cortical lesions by 7T T2-/T1-weighted MRI, and to determine relationship with other MS pathology and contribution to disability	The 7T brain scans included axial 3D MP2RAGE (0.5 mm isometric, acquired four times per scan session), sagittal 3D segmented T2w echo-planar imaging (EPI; 0.5 mm isometric, acquired in two partially overlapping volumes for full brain coverage), and axial 2D T2w multi-echo GRE (0.5 mm isometric, acquired in three partially overlapping volumes for near full supratentorial brain coverage). MP2RAGE data were processed into uniform denoised images (hereafter, T1w MP2RAGE) and T1 maps using manufacturer-provided software
Glioma	Prener et al[48]	To evaluate the potential clinical implication of the use of single-voxel magnetic resonance spectroscopy (MRS) at 7 T to assess metabolic information on lesions in a pilot cohort of patients with grades II and III gliomas	Seven patients and seven HCs were scanned using the semi-localization by adiabatic-selective refocusing sequence on a Philips Achieva 7T system equipped with a two-channel volume transmit head coil with a 32-channel receiver array. A 3D FLAIR sampled in the sagittal direction was acquired with the following parameters: TR = 7342 ms; TE = 348 ms; TI = 2200 ms; acquisition voxel size = 0.7 mm × 0.7 mm × 1.4 mm; FA = 75°; refocusing acquisition voxel size = 0.7 mm × 0.7 mm × 1.4 mm; refocusing angle = 30°; scan duration = 6 min 44 s. In this study, the scan was used for MRS voxel placement, although it was also acquired as part of an additional structural imaging research protocol
	Yuan et al[49]	The goal of the study was to explore the capability on preoperatively identifying IDH status of combining a convolutional neural network and a novel imaging modality, ultra-high-field 7T chemical exchange saturation transfer (CEST) imaging	All patients underwent MRI scans within a week prior to surgery. CEST MRI was performed on a 7T MRI scanner (MAGNETOM Terra; Siemens Healthineers, Erlangen, Germany) with a prototype-developed snapshot-CEST sequence based on a 3D gradient spoiled GRE readout with a single-channel transmit/32-channel receive head coil (Nova Medical, Wilminutesgton, MA, United States). The snapshot-CEST sequence parameters were TR = 3.4 ms, TE = 1.59 ms, FA = 6°, resolution = 1.6 mm × 1.6 mm × 5 mm, and GRAPPA acceleration factor = 3 with amplitudes B1 = 0.6, 0.75, and 0.9 mT. Z-spectra were sampled unevenly by 56 frequency offsets between -300 ppm and +300 ppm. The Z-spectrum data were corrected for both B0 and B1 inhomogeneities using the WASABI method and were fit pixelwise by a five-pool Lorentzian model (water, amide, aminutese, NOE, and MT) using the Levenberg–Marquardt algorithm. For CEST data co-registration, high-resolution T1 MP2RAGE (TR = 3800 ms, TI1 = 800 ms, TI2 = 2700 ms, TE = 2.29 ms, FA = 7°, and resolution = 0.7 mm isotropic) and 3D T2-SPACE (TR = 4000 ms, TE = 118 ms, and resolution = 0.67 mm isotropic) were acquired at 7 T. Routine-clinical-sequence, contrast-enhanced T1-weighted images (TR = 6.49 ms, TE = 2.9ms, FA = 8°, spatial resolution = 0.833 mm × 0.833 mm × 1 mm), were acquired at 3 T on an Ingenia MRI scanner (Koninklijke Philips N.V., Netherlands)
	Prener et al[55]	The aim of this study was to explore the difference in gross tumor volume of DLGGs delineated from 7T and 3T MRI scans	The 7T MR system was an actively shielded Philips Achieva 7T MR system (Philips Healthcare, Best, The Netherlands). Scanning protocol settings for 7T were as follows: 3D TSE-T2 (0.7 mm × 0.7 mm × 0.7 mm) (TE = 60) and 3D FLAIR (0.7 mm × 0.7 mm × 0.7 mm) (TE = 347). Scan time for 7T was as follows: 3D TSE-T2 was 10 min 30 s and 3DFLAIR was 7 min 30 s
	Weng et al[56]	The underlying prospective study aimed to compare SLOW-EPSI to established techniques at 7 T and 3 T for IDH-mutation status determination	The applied sequences were MEGA-SVS, MEGA-CSI, SLOW-EPSI, MRSI, 3D-T1-MPRAGE, 3D-T2-SPACE, and TOF-angiography. Measurements were performed on a MAGNETOM-Terra 7T MR scanner in clinical mode using a Nova 1Tx32Rx head coil. At 7 T, MEGA-SVS: TE = 75 ms, TR = 2500 ms, VOI = 30 × 30 × 30 mm3, averages = 64, and TA = 5 min 42 s. MEGA-CSI: TE = 75 ms, TR = 2900 ms, VOI = 70 mm × 70 mm × 20 mm, FOV = 200 × 200 × 20 mm3 (12 × 12 matrix), averages = 1, and TA = 8 min 10 s
	Yuan et al[50]	The study aimed to evaluate the diagnostic accuracy of combining CEST imaging and MRS for predicting glioma infiltration	CEST acquisitions were performed at 7 T on an MRI scanner (MAGNETOM Terra, Siemens Healthineers, Erlangen, Germany). A prototype snapshot-CEST (optimized single-shot GRE sequence with rectangular spiral reordering) was applied (TR = 3.4 ms, TE = 1.59 ms, FA = 6°, bandwidth = 660 Hz/pixel, grappa = 3, resolution = 1.6 mm × 1.6 mm × 5 mm)
	Voormolen et al[47]	To investigate intra- and extracranial distortions in 7T MRI scans of skull-base menigioma	The 7T scan parameters of the 3-dimensional sagittal magnetization-prepared turbo field echo sequence are as follows: FOV 256 mm × 256 mm × 200 mm (matrix 256 × 256 × 200), TI 1200 ms, ETL 256 mm3, readout 9 ms, TE 2.0 ms, bandwidth 506.3 Hz/pixel, and FA 8°. Total imaging time: 9 min and 36 s. Prior to the acquisition at 7 T, a B0 field map was acquired
Psychiatric disease	Van den Boom et al[51]	To investigate whether people can learn to dynamically control activity of the DLPFC, a region that has been shown to be important for working memory function and has been associated with various psychiatric disorders	fMRI performed using a 7T Philips Achieva system, with a 32-channel headcoil. Functional data recorded using an EPI sequence (TR/TE: 2.0 s/25 ms, FA: 70°, 39 axial slices, acquisition matrix 112 × 112, slice thickness 2.2 mm no gap, 2.19 mm in plane resolution). A T1-weighted image was acquired for anatomy (TR/TE: 7/2.76 ms; FA: 8°; resolution 0.98 mm × 0.98 mm × 1.0 mm)
	Morris et al[52]	To use ultra-high-field 7T MT MRI to localize the LC in humans with and without pathological anxiety, with 0.4 mm × 0.4 mm × 0.5 mm resolution in a feasible scan time. In addition, to apply a computational, data-driven LC localization and segmentation algorithm to delineate LC for all participants. The relationships between LC volume and trans-diagnostic measures of pathological anxiety and attentional control were subsequently examined in a dimensional approach based on the RDoC initiative, reflecting evidence that pathological anxiety is a trans-diagnostic construct	Participants were scanned using a 7T MRI scanner (Magnetom, Siemens, Erlangen, Germany) with a 32-channel head coil at the Leon and Norma Hess Center for Science and Medicine, ISMMS. Most subjects tolerated the MRI environment well. On entering the scanner, several subjects reported dizziness lasting 1-2 min, which they found tolerable. Structural T1-weighted dual-inversion magnetization prepared gradient echo (MP2RAGE) anatomical images were acquired first (TR = 4500 ms, TE = 3.37 ms, TI1 = 1000s and 3200 ms, FA = 4° and 5°, iPAT acceleration factor = 3, bandwidth = 130 Hz/pixel, 0.7 mm isotropic resolution, whole brain coverage). Second, MT-MRI data were acquired with a 3-D segmented GRE readout (turbo-FLASH; TFL) preceded by a train of 20 MT pulses of amplitude with 190 V transmit, and 7-min run time
	Zoon et al[53]	The study aimed to explore the connection between the location of active DBS contact points within STN subdivisions and the development of apathy in Parkinson's disease patients undergoing DBS	22 PD patients that underwent STN DBS between January 2019 and February 2020 were divided in an apathy and non-apathy group using the change in the Starkstein Apathy Scale after 6 mo of DBS. For both groups the location of DBS electrodes was determined based on 7T MRI subthalamic network analysis, enabling visualization of the subdivisions and their projections relative to the active contact point. MDS-UPDRS III scores were included to evaluate DBS effect
	Alper et al[54]	The objective of this study was to assess volumetric differences in hippocampal subfields between MDD patients globally and HCs as well as between a subset of treatment-resistant depression patients and HCs using automatic segmentation of hippocampal subfields software and ultra-high-field MRI	Thirty-five MDD patients and 28 HCs underwent imaging using 7T MRI. A 32-channel Nova Medical head coil was used to acquire brain images for segmentation. The 90-min imaging protocol included MP2RAGE (TR 6000 ms, TI1 1050 ms, TI2 3000 ms, TE 5.06 ms, voxel 0.70 mm × 0.70 mm × 0.70 mm) and T2 TSE (TR 9000 ms, TE 69 ms, voxel 0.45 mm × 0.45 mm × 2 mm) scans acquired at a coronal oblique oriented perpendicular to the long axis of the hippocampus
Essential tremor	Purrer et al[57]	The aim of the study was to analyze the localization of individual lesions with respect to the VIM and the cerebello-thalamic tract	MRI images were acquired using a 7T scanner (Siemens Magnetom, Erlangen, Germany) with a 32-channel head coil. A fast MP-RAGE sequence was utilized to acquire the WMn contrast with a high isotropic resolution of 0.7 mm. The sequence parameters were TR/TI = 4000/670 ms, matrix = 364 mm × 310 mm × 240 mm, parallel imaging acceleration factor = 3, and acquisition time = 10 min 24 s. In addition, a T1-weighted MP-RAGE was acquired with the same coverage and resolution, but TR/TI = 2500/1100 ms, FA = 7°, PA = 2, TA = 5 min 33 s

TOF: Time-of-flight; CoW: Circle of Willis; CVR: Cerebrovascular reactivity; FLAIR: Fluid-attenuated inversion recovery; dMRI: Diffusion Magnetic Resonance Imaging; OEF: Oxygen extraction fraction; ACZ: Acetazolamide; DTI: Diffusion tensor imaging; TN: Trigeminal neuralgia; MS: Multiple sclerosis; GRE: Gradient-recalled echo; MRI: Magnetic resonance imaging.