Search

Background Deep vein thrombosis (DVT) is a common but elusive illness that can lead to fatal pulmonary embolism and sudden death. Effective treatment of DVT requires accurate evaluation of thrombus distribution and stage. MRI is one of diagnostic imaging modalities for DVT, and two conventional methods are MPRAGE[1] and CE-MRV[2]. Recently, 3D T1-weighted variable-flipangle turbo spin-echo (SPACE) was proposed as a black-blood technique that permits more direct visualization of DVT[3]. However, signal suppression of tremendously slow venous blood flow remains a challenge for SPACE. The unsuppressed blood signal could be a confounder in thrombus detection[3]. We hypothesized that the 3D black-blood thrombus imaging (BTI) technique[4] that combines SPACE with DANTE black-blood preparation[5] (DANTE-SPACE) might address the above issue. Methods Experiment The IRB-approved study was performed on a 3T scanner (Siemens TimTrio, Germany). DANTE-SPACE was first optimized on 8 healthy subjects (4 F 4 M, age 25 ± 4) and then tested on 12 patients (6 F 6 M, age 52 ± 13) with DVT. The optimized parameters for DANTE included: FA 15°, pulse trains 175, RF gap 1 ms, gradient 20 mT/m. The parameters for SPACE included: 3D coronal imaging with a resolution of 1.1 × 1.1 × (1.1-1.3) mm (interpolated to 0.55 × 0.55 × [0.55-0.65] mm),TR/ TE 650/9.8 ms, turbo factor 40, GRAPPA 2, scan time ~4 min. The scan was targeted to the thrombus region that was pre-determined by ultrasound within 3 days. Conventional SPACE, MPRAGE, and CE-MRV were conducted for comparison.

Background Mouse models with mycardial infarction (MI) have been intensively used to investigate the cardiac remodeling and functional change [1,2]. MRI with ECG triggering and respiratory gating (or breath-holding) is commonly used to assess MI due to its noninvasive nature. However, the manipulations of external ECG triggering and respiratory gating are cumbersome due to the small size of mouse. To address this issue, a 3D self-gating (SG) MR technique with stack-of-stars sampling trajectories was proposed for retrospectively cardiac and respiratory-gated MI imaging in mouse models [3]. Methods: Technical Design: MR data was acquired by a T1-weighted GRE sequence with stack-of-stars sampling trajectories and a partition-first golden-angle reordering (Figure 1a). The centers of k-space lines acquired at the same angle from individual partitions were aggregated and used as an SG time point. The periodic respiratory and cardiac motions were then detected through an iterative filtering process on the SG time series according to the cardiac rate of 300~500 per minute and respiration rate of 70~110 per minute. After resorting the imaging data into appropriate cardiac and respiratory phases, motion-artifact-free were finally reconstructed. Experiments: The SG technique was preliminarily validated on 5 mice with MI induction and all MR scans were performed on a 3T scanner (Siemens Tim Trio, Germany) with a customized 4-channal mouse coil. Typical imaging parameters for the ungated GRE sequence included: flip angle = 18°, TR = 4.2 ms, TE = 2.4 ms, spatial resolution = 0.6 × 0.6 × 1.5 mm, bandwidth = 620 Hz/Pixel, partition number = 12, and a total number of 3200 projections were continuously collected, corresponding to a fixed scan time of 3 min. An amount of 0.5 ml gadolinium contrast agent with concentration of 0.5 mmol/ml (Consun Pharmaceutical Group Limited, GuangZhou, China) was injected to enhance the MI. The mice were sacrificed immediately after MRI for histological analysis and comparison to the MR results.

Methods CMR perfusion imaging was performed in six canines and thirty patients at 1.5T using dual-bolus (0.005 and 0.05 mmol/kg Gd-DTPA) and dual-sequence techniques with 1RR, 90° composite pulse, 50° SSFP readout, saturation recovery 90 ms, TR 2.4 ms, TE 1.2 ms, matrix size 128 × 80. A low TE 0.6 ms, low-resolution 64 × 48 FLASH image series was also acquired. The AIF was measured from the low-dose high-resolution series (DB), the high-dose low-resolution series (DS), and the high-dose high-resolution conventional single-bolus series (SB). Myocardial time intensity curves were analyzed on a midslice based on 6 transmural sectors and quantified by model-constrained deconvolution.

To assess the global and segmental left ventricular (LV) native T1 and extracellular volume fraction (ECV) in children and young adults with hypertrophic cardiomyopathy (HCM) compared to a control cohort. The study population included 21 HCM patients (mean 14.1 ± 4.6 years) and 21 controls (mean 15.7 ± 1.5 years). Native modified Look-Locker inversion recovery sequence was performed before and after contrast injection in 3 short axis planes. Global and segmental LV native T1 and ECV were quantified and compared between HCM patients and controls. Mean native T1 in HCM patients and controls was 1020.4 ± 41.2 and 965.6 ± 30.2 ms respectively (p < 0.0001). Hypertrophied myocardium had significantly higher native global T1 and global ECV compared to non-hypertrophied myocardium in HCM (p < 0.0001, = 0.14 and 0.048, = 0.01 respectively). In a subset of patients, ECV was higher in LV segments with LGE compared to no LGE (p < 0.0001). No significant correlation was identified between global native T1 and ECV and parameters of LV structure and function. Native T1 cut-off of 987 ms provided the highest sensitivity (95 %) and specificity (91 %) to separate HCM patients from controls. Global and segmental native T1 are elevated in HCM patients. LV segments with hypertrophy and/or LGE had higher ECV in a subset of HCM patients. LV native T1 and ECV do not correlate with parameters of LV structure and function. T1 in children and young adults may be used as a non-invasive tool to assess for HCM and related fibrosis.

International audience

Background Hemorrhage is a frequent complication in acute myocardial infarction (AMI) and is speculated to be an independent predictor of adverse outcomes [1]. In addition to infarct zone remodeling, the distal remote myocardium may experience alterations vasodilator dysfunction, edema, and extracellular matrix (ECM) expansion [2-4]. The aim of our study was to understand the impact of hemorrhage on infarct as well as remote myocardial remodeling following an ischemic event. To this end, we employed quantitative T1 and T2 mapping to probe the underlying tissue alterations in a novel hemorrhagic model of AMI.

Background Cardiac phase-resolved Blood-Oxygen-Level-Dependent (CP-BOLD) MRI can detect myocardial ischemia at rest without contrast and stress agents. At rest, BOLD myocardial signal intensity varies with cardiac phase: in healthy conditions it is maximal in systole and minimal in diastole, but in disease this pattern is not evident. These changes are not readily visible and post-processing is necessary. Previous methods used segmental analysis and only two images to identify ischemic segments. In this study we demonstrate that it is possible to use unsupervised learning methods to identify ischemia with higher accuracy.

Background CMR can be used to assess the structure and function of athletes’ hearts. However, diastolic function is not routinely measured with CMR. Mitral annular (MA) motion, commonly assessed by tissue Doppler imaging, is an important element of diastole that may help to distinguish physiological remodeling from pathological processes of the heart. CMR has been shown to play a useful role in evaluating MA motion. This study aims to characterize the global left ventricular diastolic function in endurance athletes by 3D volume tracking of the MA, using conventional cine-CMR images. Methods CMR studies of 24 high-level endurance athletes (age 22 ± 3) were retrospectively selected from a database. Routine CMR studies from 39 patients with normal hearts (age 42 ± 16) and 29 hypertrophic cardiomyopathy (HCM) patients (age 52 ± 17) with normal ejection fractions were included for comparison. Using normalized cross-correlation, feature-tracking of the atrioventricular junctions was performed over the cardiac cycle, in routine long-axis 2-, 3-, and 4-chamber cine-CMR views. This resulted in 6 spatial MA points per cardiac phase, which were interpolated with a spline for 3D MA reconstruction. The 3D volume swept out by the MA was then calculated; the net sweep volume at a given cardiac phase was derived from the sum of the incremental volumes starting from end-diastole.

Background Echocardiographic tissue Doppler or magnetic resonance imaging (MRI) measurements of early diastolic mitral annular velocity with other appropriate parameters are frequently used as a non-invasive diagnostic tool of diastolic dysfunction. Previously published global approach utilizing normalized left ventricular (LV) torsion shear angle volume loop ( _hat V_hat loop) was proposed to provide a new global description of LV diastolic function. The purpose of this study was to evaluate the discriminant power of these non-invasive parameters in identifying elevated LV end-diastolic pressure (LVEDP) (i.e. LVEDP≥15ml) and provide a non-invasive index to predict elevated LVEDP. Methods A group of 23 patients with LV ejection fraction≥50% without acute infarct undergoing cardiac catheterization that did not undergo percutaneous coronary intervention were studied using high-fidelity pressure measurement. Echocardiogram with flow and tissue Doppler quantification was performed on the same date of cardiac catheterization for all participants. Cine and tagged cardiac MRI were performed on all subjects, followed by comprehensive volumetric and strain analysis. The database consisted of five parameters collected from all subjects: torsion hysteresis area (THA), peak -d_hat/dV_hat at early diastole, MRI derived E/AMRI, echocardiographic derived E/A and E/e’. Stepwise variable selection was applied to select parameters with significance level of leave out and stay in equal to 0.15. A logistic regression classifier was used to construct the non-invasive index for identifying elevated LVEDP based on the selected parameters. The classifier’s prediction performance was analyzed using a Receiver-Operating Characteristic (ROC) curve and expressed as its sensitivity, specificity, accuracy and area under the curve (AUC). Results

Background An understanding of inflammation following myocardial infarction may be of importance in the development of novel therapeutics to limit the development of heart failure following myocardial injury. However, the quantification of inflammation in this setting continues to be challenging. Hybrid FDG PET/MR may offer a non-invasive in vivo solution through intrinsic registration of these complementary modalities. This study sought to validate its use in a canine model of myocardial infarction (MI).

Background The clinical translation of MRI-guided cardiovascular catheterization has been limited by the unavailability of devices that are both visible and safe under MRI. In particular, rigid metallic guidewires are essential for most catheterization procedures and are at risk of heating during MR imaging [1]. Here we present an MRI method that simultaneously improves the visualization of commercially available nitinol guidewires and minimizes RF induced heating.

Background Conventional, planar phase-contrast (2D-PC) imaging is the gold standard for non-invasive measurement of blood flow, routinely used in the assessment of structural heart disease by MRI. Nevertheless, at many institutions, nuclear perfusion scintigraphy (NPS) remains necessary for confirmation of differential pulmonary perfusion, but requires an additional exam with radiation exposure, and in younger children prolongs cardiac anesthesia. Highly-accelerated 4D phase-contrast

Background Blood flow quantification using ECG gated phase contrast (PC) MRI has proven to be a useful clinical tool for the evaluation of cardiovascular pathologies such aortic valve disease or cardiac shunts. However, data are acquired in an ECG synchronized manner over multiple heart beats and the impact of cardiac arrhythmia, which can frequently be encountered in patients with atrial fibrillation (AF), on flow parameters is unclear. To assess the impact of arrhythmia and thus RR interval variability on ECG-gated PC MRI acquisitions we have performed a simulation study to systematically investigate the impact of beat-to-beat variations on ECG gated multi-beat flow imaging with MRI using real time invivo TEE data in 5 AF patients with known arrhythmia.

Background Obesity is a risk factor for cardiovascular disease and mortality. Studies in both obese humans and murine models of obesity have identified changes in left ventricular (LV) mechanics (i.e., strains, strain rates, and torsion), which manifest prior to global changes in cardiac function (ejection fraction) and may represent early markers of cardiovascular disease. These data are generally acquired under resting conditions, which could mask subtle differences in the early stages of disease. We sought to evaluate LV mechanics under inotropic stress conditions with the hypothesis that mechanical deficiencies with obesity would be exacerbated under stress conditions and revealed at earlier stages of disease.

Background Cardiotoxicity is a major limiting factor preventing dose-effective use of multiple cytotoxic agents in cancer treatment. Current cardiotoxicity diagnosis is based on functional changes that occur late in the course of the disease. In this pre-clinical pilot study, we hypothesize myocardial T1 and T2 mapping will identify mice at risk of death secondary to doxorubicin-induced cardiotoxicity prior to cardiac function deterioration. Methods Eleven healthy, 10-week-old C57BL/6 female mice were included in the study. All procedures are in accordance with an Institutional Animal Care and Use Committee approved protocol. Doxorubicin-induced cardiotoxicity was generated by giving 8 mg/kg of doxorubicin via intraperitoneal (IP) injection weekly for 4 weeks in n = 8 randomly selected mice. A control group of n = 3 mice was given a weekly volume equivalent dose of IP normal saline. Each mouse underwent ECG-gated CMR on a 7T small animal scanner (ClinScan, Bruker, Billerica, MA) prior to treatment, at 2 weeks, and post-treatment. Steady-state free precession images were acquired at the left ventricular base, mid-ventricle, and apex and ejection fraction (EF) was calculated (Figure 1). Gadopentetate Dimeglumine contrast was injected via the tail vein at a dose of 1 mL/kg. Myocardial T1 and T2 maps were generated using dedicated offline software to calculate the pixel-wise relaxation

Background Conventional breath-held gadolinium-based contrastenhanced MR angiography (CE-MRA) provides excellent definition of extra-cardiac anatomy, but does not provide diagnostic image quality for intra-cardiac anatomy because it is not gated to ECG due to time constraints associated with breath-holding and the need to capture the first-pass of gadolinium. To address this issue, we propose a 4D non-breath-held multiphase steady-state imaging sequence (MUSIC) using ferumoxytol, which is an FDA approved iron-oxide particle for treating irondeficiency anemia, as an intravascular contrast agent and demonstrate its feasibility in children with congenital heart disease (CHD).

Background Displacement Encoding with Stimulated Echoes (DENSE) directly measures tissue displacements and can be used to quantify cardiac mechanics. Multi-dimensional DENSE results in lengthy scans that require respiratory gating, acquiring data only while the diaphragm is within a prespecified “acceptance window.” Because it is not possible to perform respiratory gating during data acquisition, DENSE can employ the following respiratory gating strategies: 1) acquire data and keep it if the diaphragm is inside the window after acquisition (retrospective) 2) acquire data and keep it only if the diaphragm was inside the window right before data acquisition (prospective) or 3) a combination of retrospective and prospective where the diaphragm must be inside the window both before and after data acquisition (combined) (Fig 1). Combined respiratory gating is not used often because more data is discarded, resulting in longer scan times. It is possible, however, that with only retrospective or prospective respiratory gating, the diaphragm may not be within the acceptance window for the entirety of data acquisition (i.e. drifting into or out of the window), negatively affecting image quality. We hypothesized that the combined respiratory gating would result in significantly different estimates of left ventricular strains compared to either retrospective or prospective respiratory gating.

Methods 12 healthy controls (age 9-53 years) and 8 patients (age 8-37 years) with a corrected atrioventricular septal defect and a mean mitral valve regurgitation of 25% (range 1937%) were included. The maximal left atrial volume (LAV) was calculated based on the biplane area-length method from a standardized 2and 4-chamber view. Whole-heart 4D Flow MRI was performed on a 3T MR scanner with free breathing, three-directional velocity encoding of 150cm/s in all directions, spatial resolution 2.3×2.3×3.0-4.2mm and 30 phases reconstructed over one cardiac cycle. At end-systole the vortical flow was assessed by manually segmenting the volume of circular flow in the left atrium based on streamline visualization in a stack of slices parallel to the 4-chamber view. 3Dparticle tracing was applied in reverse time order, using the defined volume as seeding, to trace back the seed points and quantify the number of particles originating from the left and right pulmonary veins (LPV and RPV). Results In controls mean LAV was 60±35mL and the vortex flow volume at end-systole 8±5mL. Tracing revealed a dominating contribution to the vortical volume originating from the LPV (41±14%), a smaller part from the RPV (17±12%) and a residual part of particles present inside the atrium at the start of systole (42±15%). In patients with mitral regurgitation the LAV was 76 ±21mL and a complex shape of the vortical flow with variation between subjects was observed (Figure 1). Mean vortical volume at end-systole was 5±4mL, with 13±14% contribution from LPV (difference with controls, p=0.001), 6±6% from the RPV (p=0.049) and a residual part of 81 ±14% (particles already inside the atrium at start of systole or coming from the ventricle as regurgitation) (p

Background The effects of ischemia are first realized in the subendocardium and progress transmurally. Hence, detection of the transmural extent of myocardial perfusion (MP) defects has important diagnostic/prognostic implications. Systolic MP imaging using ECG-gated methods has been shown to provide improved visualization of subendocardial defects [1]. We developed an innovative non-ECGgated perfusion CMR technique capable of imaging all slices at the same systolic phase, and hypothesized that it improves visualization of the transmural extent of MP defects compared to the conventional method in CMR studies of CAD patients.

BACKGROUND Bicuspid aortic valve (BAV)-related aortopathy is characterized by histological abnormalities that result in aortic wall stiffening and aortic dilation. The study aim was to determine the range of ascending aortic stiffness seen in a clinical cohort of patients with BAV, and to identify the association of aortic stiffness with standard clinical and imaging parameters. METHODS Patients with BAV (n = 65) and normal subjects (n = 10) were studied using conventional phase-contrast magnetic resonance imaging through the ascending aorta. Local aortic stiffness was estimated by measuring pulse wave velocity (PWV) using the flow-area (QA) method. Correlations between PWV and other imaging and clinical variables were assessed, and multiple linear regression models were used to examine independent predictors of PWV. RESULTS BAV patients demonstrated a significantly higher mean PWV compared to normal subjects (6.53 ± 5.88 versus 3.51 ± 0.92 m/s; p

Background Aortic Coarctation (AoCo) accounts for 5-8% of the children with CHD. Even after successful early repair, life expectancy is still markedly reduced (80% at 50 years after surgery) compared to normal population due to long term complications (hypertension). Usually, invasive diagnostic catheter investigations are required to evaluate the pressure gradient across the aorta at rest, or unmask such a gradient by use of isoprenaline stress to mimic physical exercise. The application of imagebased computational fluid dynamics (CFD) in patients with AoCo appears promising as an alternative noninvasive diagnostic tool, as it may allow the avoidance of cardiac catheterization to determine pressure gradients. The motivation of this research is to know if a MRI based CFD model can accurately predict the pressure

Background The homogeneous distribution of collagen in diffuse myocardial fibrosis renders the disease unsuitable for imaging using late gadolinium enhancement (LGE) [1]. More recently, the estimation of extracellular volume from T1 maps involving gadolinium agents has shown promise; however, these methods are not specific to collagen and are governed by gadolinium kinetics [2]. The diagnosis of diffuse myocardial fibrosis would benefit from an imaging method that can directly detect collagen. Notably, ultra-short echo time magnetic resonance (UTE MR) is a technique that can be used to detect short T2* species, including collagen [3]. Our objective is to characterize the UTE signal of protons in the collagen molecule, including their T2* and chemical shift. Direct isolation of a collagen signal could aid in the diagnosis of myocardial fibrosis, especially for diffuse distributions, and the assessment of disease extent.

Background Human physiology and disease are defined by complex interactions between genetic and environmental factors. While genome-wide association studies (GWASs) have been successfully used to dissect many human traits, they have not worked for studies of the heart. This likely reflects the limitations in echocardiographic and traditional magnetic resonance phenotyping, as many cardiac traits are known to be highly heritable. In an attempt to overcome the limitations of human heart GWASs to date, we developed high-resolution 3D-cardiac MRI (3D-CMR) techniques combined with computational image analysis and dense genotyping to examine the common genetic determinants of cardiac structure and function.

Background Myocardial oedema (area-at-risk, AAR) is typically imaged using a pre-contrast T2-weighted short tau inversion recovery (T2w-STIR) sequence on cardiovascular magnetic resonance (CMR) imaging. However, this sequence is prone to motion and rhythm artefact, signal dropout, blood-pool artefact, surface coil signal inhomogeneity and potentially prohibitive long breath-hold duration. This susceptibility to artefacts limits utility of T2w-STIR in large clinical trials where attainment of diagnostic quality oedema imaging in the majority is necessary to determine myocardial salvage: a measure of reperfusion success and a strong predictor of adverse remodeling and prognosis post ST-segment elevation myocardial infarction (STEMI). We compare AAR quantified on T2w-STIR imaging with novel T1-mapping on 3.0T CMR post STEMI.

Background Conventional inversion recovery (IR) techniques for the detection of late gadolinium enhancement (LGE) in the myocardium are bright blood methods in which the signal in the blood chamber is often similar to the signal in hyperenhanced irreversibly damaged myocardial areas. Due to the low contrast to noise ratio (CNR) between blood and the hyperenhanced myocardium at the tissue-blood interface, the discrimination of subendocardial LGE from blood is often challenging with this approach. The aim of our study was to evaluate the accuracy of a prototype dark blood LGE technique for the detection and quantification of myocardial LGE in patients with myocardial infarction (MI) when compared to conventional IR LGE technique.

Background Displacement Encoding with Stimulated Echoes (DENSE) encodes displacement into the phase of the magnetic resonance signal. Due to the stimulated echo, the signal is inherently low and fades through the cardiac cycle. To compensate, a spiral acquisition has been used at 1.5T. This spiral sequence has not been validated at 3T, where the increased signal would be valuable, but field inhomogeneities may result in measurement errors. We hypothesized that spiral cine DENSE is valid at 3T and tested this hypothesis by measuring displacement errors at both 1.5T and 3T in vivo. Methods Two-dimensional spiral cine DENSE and tagged imaging of the left ventricle were performed on ten healthy subjects at 3T and six healthy subjects at 1.5T. Intersection points were identified on tagged images near end-systole. Displacements from the DENSE images were used to project those points back to their origins. The deviation from a perfect grid was used as a measure of accuracy and quantified as root-mean-squared error. This measure was compared between 3T and 1.5T with the Wilcoxon rank sum test. Inter-observer variability of strains and torsion quantified by DENSE and agreement between DENSE and harmonic phase (HARP) were assessed by Bland-Altman analyses. The signal to noise ratio (SNR) at each cardiac phase was compared between 3T and 1.5T with the Wilcoxon rank sum test. Results The displacement accuracy of spiral cine DENSE was not different between 3T and 1.5T (1.2 ± 0.3 mm and 1.2 ± 0.4 mm, respectively). Both values were lower than the DENSE pixel spacing of 2.8 mm. There were no substantial differences in inter-observer variability of DENSE or agreement of DENSE and HARP between 3T and 1.5T. Relative to 1.5T, the SNR at 3T was greater by a factor of 1.4 ± 0.3. Conclusions The spiral cine DENSE acquisition that has been used at 1.5T to measure cardiac displacements can be applied at 3T with equivalent accuracy. The inter-observer variability and agreement of DENSE-derived peak strains and torsion with HARP is also comparable at both field strengths. Future studies with spiral cine DENSE may take advantage of the additional SNR at 3T.

Background Late gadolinium enhancement (LGE) imaging is the gold standard for noninvasive evaluation of myocardial scar. In the clinical setting, the standard imaging protocol involves a breath-held segmented inversion-prepared 2D Cartesian acquisition at a single location. This breathhold scan is repeated 10 -14 times to cover the entire left ventricle. Although multislice 2D LGE has shown great diagnostic accuracy, the multiple breath-holds increase the scan time to approximately 10min and patient fatigue which can result in poor breath-holding and ghosting artifact. In this study, we propose to perform the 3D LGE imaging in a single breath-hold using a stack of spiral trajectory. Methods Single breath-hold 3D stack of spiral LGE images covering the entire LV were acquired in 27 subjects undergoing clinical scans following the conventional multisilce 2D LGE on a Siemens 1.5T Avanto scanner. The 3D spiral LGE sequence consisted of 12 partitions of a dual density spiral trajectory. Each spiral readout was 4ms long with 24 interleaves to support 1.5x Nyquist in the center and 0.7x Nyquist in the edge of kspace. At each cardiac cycle, 2 out of 24 interleaves were acquired for each partition resulting in an acquisition window of approximately 160ms. All of the spirals were acquired in a single 12 heart beat breath-hold. Other sequence parameters included: TR 7ms, TE 1ms, TI 300~400ms, FA 20, FOV 340mm, in-plane resolution 1.5mm, 12 slices, and thickness 8mm. The images

Background Blood flow in the left ventricle (LV) is closely linked to the function of valves, great vessels and the myocardium. Previous studies have used the Pressure Poisson Equation (PPE) to compute relative pressure fields from 4D flow data. However, the PPE may be numerically sensitive to errors in velocities and delineations. Hemodynamic forces is a quantitative measure similar to relative pressure maps, which may be less sensitive to errors. Therefore, the aim of this study was to investigate the reproducibility of hemodynamic force quantification, and to present initial observations in controls, elite endurance athletes and patients with heart failure.