Wouter Koning

Current Status of Clinical Magnetic Resonance Imaging for Plaque Characterisation in Patients with Carotid Artery Stenosis

OBJECTIVE The article aims to provide an overview of the literature that assessed the agreement between magnetic resonance imaging (MRI) and histology for specific carotid plaque characteristics associated with vulnerability in terms of sensitivity and specificity. METHODS A systematic search strategy was conducted in MEDLINE and EMBASE databases resulting in 1084 articles. Finally, we included 17 papers. Due to variation in presentation, especially in MRI and histology methods, a pooled analysis could not be performed. RESULTS Two studies were performed on a 3.0-T MRI scanner; all other studies were performed on a 1.5-T scanner. Most performed sequences were two-dimensional (2D) and three-dimensional (3D) T1-weighted and all histology protocols varied slightly. Our results indicate that calcification, fibrous cap, intraplaque haemorrhage and lipid-rich necrotic cores can be identified with moderate-to-good sensitivity and specificity. CONCLUSIONS Based on current literature, it appears premature for routine application of MRI as an imaging modality to assess carotid plaque characteristics associated with plaque vulnerability. Although MRI still holds promise, clinical application for plaque characterisation would require consensus regarding MRI settings and confirmation by histology. Predefined protocols for histology and MR imaging need to be established.

High-resolution MRI of the carotid arteries using a leaky waveguide transmitter and a high-density receive array at 7 T

A setup for 7T MRI of the carotid arteries in the neck was designed and constructed. Separate dedicated arrays were used for transmit and receive. For the transmit array, single‐side adapted dipole antennas were mounted on a dielectric pillow, which was shown to serve as a leaky waveguide, efficiently distributing B1 into the neck. Risk assessment was performed by simulations. Phantom measurements were performed to establish optimal positions of the antennas on the pillow. Using two antennas, a dual transmit setup was created. In vivo B1+ maps with different shim configurations were acquired to assess transmit performance. This effective transmit array was used in combination with a dedicated 30 channel small element receive coil. High‐resolution in vivo turbo spin echo images were acquired to demonstrate the excellent performance of the setup. Magn Reson Med 69:1186–1193, 2013.

MRI of the carotid artery at 7 Tesla: quantitative comparison with 3 Tesla.

To evaluate the 7 Tesla (T) MRI of the carotid arteries, as quantitatively compared with 3T.

Seven-Tesla Magnetic Resonance Imaging of Atherosclerotic Plaque in the Significantly Stenosed Carotid Artery A Feasibility Study

ObjectivesThe objective of this study was to assess the feasibility of carotid vessel wall imaging at 7.0 for T magnetic resonance imaging (MRI) in a series of patients with a symptomatic greater than 70% stenosis of the internal carotid artery. Materials and MethodsFirst, a series of 6 healthy volunteers were scanned at 3.0 T and 7.0 T MRI to perform a signal-to-noise ratio comparison between these 2 field strengths. Second, in patients with a greater than 70% stenosed carotid artery, a 7.0 T MRI protocol, consisting of a dual-echo turbo spin echo sequence (echo times of 45 and 150 milliseconds) and a T1-weighted turbo spin echo sequence, was obtained. Lumen and vessel wall were delineated for interobserver and intraobserver reproducibility, and signal intensity distribution in the most severely stenosed part of the internal carotid artery was correlated with different plaque components on histopathologic findings. ResultsThe mean (SD) signal-to-noise ratio in the vessel wall was 42 (12) at 7.0 T and 24 (4) at 3.0 T. Nineteen patients were included, but technical issues yielded carotid MRI data of 14 patients available for the final analysis. Of these patients, 4 were diagnosed with stroke, 7 were diagnosed with a transient ischemic attack, and 3 were diagnosed with amaurosis fugax. Intraclass correlation coefficient of the agreements of lumen and vessel wall determination between 2 observers and between the repeated measures of 1 observer were above 0.80 in both 3.0 T and 7.0 T data sets of the healthy volunteers and also in the 7.0 T data set of the patients. Signal hyperintensity in the 7.0 T magnetic resonance images was inversely proportional to calcification. Other correlations between plaque components and signal intensity could not be confirmed. ConclusionsThis first series of patients with carotid atherosclerotic plaque who were scanned at 7.0 T MRI shows that 7.0 T MRI enables to adequately determine lumen and vessel wall areas. Signal hyperintensity in these 7.0 T magnetic resonance images was inversely proportional to calcification. However, at this stage, no other correlations between histologic findings and vessel wall contrast were found. Implementation of in vivo high-resolution 7.0 T MRI of plaque components for risk stratification remains challenging. Future development of hardware and software is still needed to attain a more robust setup and to enable complete plaque characterization, similar to what is currently possible with multiple MRI sequences at 1.5 T and 3.0 T MRI.

Adiabatic turbo spin echo in human applications at 7 T

Nonuniform B1 fields in ultrahigh‐field MR imaging cause severe image artifacts, when conventional radiofrequency (RF) pulses are used. Particularly in MR sequences that encompass multiple RF pulses, e.g., turbo spin echo (TSE) sequences, complete signal loss may occur in certain areas. When using a surface coil for transmitting the RF pulses, these problems become even more challenging, as the spatial B1 field variance is substantial. As an alternative to conventional TSE sequences, adiabatic TSE sequences can be applied, which have the benefit that these sequences are insensitive to B1 nonuniformity. In this study, we investigate the potential of using adiabatic TSE at 7 T with surface coil transceivers in human applications. The adiabatic RF pulses were tuned to deal with the constraints in B1 strength and RF power deposition, but remained in the superadiabatic regime. As a consequence, the dynamic range in B1 is compromised, and signal modulation is obtained over the echo train. Multidimensional Bloch simulations over the echo train and phantom measurements were obtained to assess these limitations. Still, using proper k‐space sampling, we demonstrate improved image quality of the adiabatic TSE versus conventional TSE in the brain, the neck (carotid artery) and in the pelvis (prostate) at 7 T. Magn Reson Med, 2012.

7.0 T MRI detection of cerebral microinfarcts in patients with a symptomatic high-grade carotid artery stenosis.

In the current study, the presence of cerebral cortical microinfarcts (CMIs) was evaluated in a series of 21 patients with a symptomatic high-grade > 50% stenosis of the carotid artery. A T2-weighted fluid-attenuated inversion recovery sequence and a T1-weighted turbo field echo sequence of the brain were obtained at 7.0 Tesla magnetic resonance imaging. Primary study endpoint was the number of CMIs and macroinfarcts. In total, 53 cerebral infarcts (35 macroinfarcts; 18 CMIs) were found ipsilateral to the symptomatic carotid artery, in 14 patients (67%). In four of these patients, both CMIs and macroinfarcts were visible. In the contralateral hemisphere, seven infarcts (five macroinfarcts and two CMIs) were found in five patients (24%). In the ipsilateral hemispheres, the number of CMIs and macroinfarcts were significantly correlated (P = 0.02). Unpaired comparison of medians showed that the number of CMIs in the ipsilateral hemisphere was significantly higher than the number of CMIs in the contralateral hemisphere (P = 0.04). No significant correlation was found between stenosis grade and the number of any infarct. The current study shows that in symptomatic patients with significant extracranial carotid artery stenosis, CMIs are part of the total cerebrovascular burden and these CMIs prevail with a similar pattern as observed macroinfarcts.

PLACD-7T Study: Atherosclerotic Carotid Plaque Components Correlated with Cerebral Damage at 7 Tesla Magnetic Resonance Imaging

Introduction: In patients with carotid artery stenosis histological plaque composition is associated with plaque stability and with presenting symptomatology. Preferentially, plaque vulnerability should be taken into account in pre-operative work-up of patients with severe carotid artery stenosis. However, currently no appropriate and conclusive (non-) invasive technique to differentiate between the high and low risk carotid artery plaque in vivo is available. We propose that 7 Tesla human high resolution MRI scanning will visualize carotid plaque characteristics more precisely and will enable correlation of these specific components with cerebral damage. Study objective: The aim of the PlaCD-7T study is 1: to correlate 7T imaging with carotid plaque histology (gold standard); and 2: to correlate plaque characteristics with cerebral damage ((clinically silent) cerebral (micro) infarcts or bleeds) on 7 Tesla high resolution (HR) MRI. Design: We propose a single center prospective study for either symptomatic or asymptomatic patients with haemodynamic significant (70%) stenosis of at least one of the carotid arteries. The Athero-Express (AE) biobank histological analysis will be derived according to standard protocol. Patients included in the AE and our prospective study will undergo a pre-operative 7 Tesla HR-MRI scan of both the head and neck area. Discussion: We hypothesize that the 7 Tesla MRI scanner will allow early identification of high risk carotid plaques being associated with micro infarcted cerebral areas, and will thus be able to identify patients with a high risk of periprocedural stroke, by identification of surrogate measures of increased cardiovascular risk.

Dielectric waveguides for ultrahigh field magnetic resonance imaging.

The design of RF coils for MRI transmit becomes increasingly challenging at high frequencies required for MRI at 7T and above. Our goal is to show a proof of principle of a new type of transmit coil for higher field strengths.

(1) H MRS in the human spinal cord at 7 T using a dielectric waveguide transmitter, RF shimming and a high density receive array

Multimodal MRI is the state of the art method for clinical diagnostics and therapy monitoring of the spinal cord, with MRS being an emerging modality that has the potential to detect relevant changes of the spinal cord tissue at an earlier stage and to enhance specificity. Methodological challenges related to the small dimensions and deep location of the human spinal cord inside the human body, field fluctuations due to respiratory motion, susceptibility differences to adjacent tissue such as vertebras and pulsatile flow of the cerebrospinal fluid hinder the clinical application of 1H MRS to the human spinal cord. Complementary to previous studies that partly addressed these problems, this work aims at enhancing the signal‐to‐noise ratio (SNR) of 1H MRS in the human spinal cord. To this end a flexible tight fit high density receiver array and ultra‐high field strength (7 T) were combined. A dielectric waveguide and dipole antenna transmission coil allowed for dual channel RF shimming, focusing the RF field in the spinal cord, and an inner‐volume saturated semi‐LASER sequence was used for robust localization in the presence of B1+ inhomogeneity. Herein we report the first 7 T spinal cord 1H MR spectra, which were obtained in seven independent measurements of 128 averages each in three healthy volunteers. The spectra exhibit high quality (full width at half maximum 0.09 ppm, SNR 7.6) and absence of artifacts and allow for reliable quantification of N‐acetyl aspartate (NAA) (NAA/Cr (creatine) 1.31 ± 0.20; Cramér–Rao lower bound (CRLB) 5), total choline containing compounds (Cho) (Cho/Cr 0.32 ± 0.07; CRLB 7), Cr (CRLB 5) and myo‐inositol (mI) (mI/Cr 1.08 ± 0.22; CRLB 6) in 7.5 min in the human cervical spinal cord. Thus metabolic information from the spinal cord can be obtained in clinically feasible scan times at 7 T, and its benefit for clinical decision making in spinal cord disorders will be investigated in the future using the presented methodology. Copyright