Simultaneous left atrium anatomy and scar segmentations via deep learning in multiview information with attention.

Three-dimensional late gadolinium enhanced (LGE) cardiac MR (CMR) of left atrial scar in patients with atrial fibrillation (AF) has recently emerged as a promising technique to stratify patients, to guide ablation therapy and to predict treatment success. This requires a segmentation of the high intensity scar tissue and also a segmentation of the left atrium (LA) anatomy, the latter usually being derived from a separate bright-blood acquisition. Performing both segmentations automatically from a single 3D LGE CMR acquisition would eliminate the need for an additional acquisition and avoid subsequent registration issues. In this paper, we propose a joint segmentation method based on multiview two-task (MVTT) recursive attention model working directly on 3D LGE CMR images to segment the LA (and proximal pulmonary veins) and to delineate the scar on the same dataset. Using our MVTT recursive attention model, both the LA anatomy and scar can be segmented accurately (mean Dice score of 93% for the LA anatomy and 87% for the scar segmentations) and efficiently ( ∼ 0.27 s to simultaneously segment the LA anatomy and scars directly from the 3D LGE CMR dataset with 60-68 2D slices). Compared to conventional unsupervised learning and other state-of-the-art deep learning based methods, the proposed MVTT model achieved excellent results, leading to an automatic generation of a patient-specific anatomical model combined with scar segmentation for patients in AF.

Deep Learning for Diagnosis of Chronic Myocardial Infarction on Nonenhanced Cardiac Cine MRI.

Background Renal impairment is common in patients with coronary artery disease and, if severe, late gadolinium enhancement (LGE) imaging for myocardial infarction (MI) evaluation cannot be performed. Purpose To develop a fully automatic framework for chronic MI delineation via deep learning on non-contrast material-enhanced cardiac cine MRI. Materials and Methods In this retrospective single-center study, a deep learning model was developed to extract motion features from the left ventricle and delineate MI regions on nonenhanced cardiac cine MRI collected between October 2015 and March 2017. Patients with chronic MI, as well as healthy control patients, had both nonenhanced cardiac cine (25 phases per cardiac cycle) and LGE MRI examinations. Eighty percent of MRI examinations were used for the training data set and 20% for the independent testing data set. Chronic MI regions on LGE MRI were defined as ground truth. Diagnostic performance was assessed by analysis of the area under the receiver operating characteristic curve (AUC). MI area and MI area percentage from nonenhanced cardiac cine and LGE MRI were compared by using the Pearson correlation, paired t test, and Bland-Altman analysis. Results Study participants included 212 patients with chronic MI (men, 171; age, 57.2 years ± 12.5) and 87 healthy control patients (men, 42; age, 43.3 years ± 15.5). Using the full cardiac cine MRI, the per-segment sensitivity and specificity for detecting chronic MI in the independent test set was 89.8% and 99.1%, respectively, with an AUC of 0.94. There were no differences between nonenhanced cardiac cine and LGE MRI analyses in number of MI segments (114 vs 127, respectively; P = .38), per-patient MI area (6.2 cm2 ± 2.8 vs 5.5 cm2 ± 2.3, respectively; P = .27; correlation coefficient, r = 0.88), and MI area percentage (21.5% ± 17.3 vs 18.5% ± 15.4; P = .17; correlation coefficient, r = 0.89). Conclusion The proposed deep learning framework on nonenhanced cardiac cine MRI enables the confirmation (presence), detection (position), and delineation (transmurality and size) of chronic myocardial infarction. However, future larger-scale multicenter studies are required for a full validation. Published under a CC BY 4.0 license. Online supplemental material is available for this article. See also the editorial by Leiner in this issue.

Pulmonary artery diameters, cross sectional areas and area changes measured by cine cardiovascular magnetic resonance in healthy volunteers.

BACKGROUND: We measured by cine cardiovascular magnetic resonance (CMR) main and branch pulmonary artery diameters and cross sectional areas in diastole and systole in order to establish normal ranges and the effects on them of age, gender and body surface area (BSA). Documentation of normal ranges provides a reference for research and clinical investigation in the fields of congenital heart disease, pulmonary hypertension and connective tissue disorders. METHODS: We recruited 120 healthy volunteers: ten males (M) and ten females (F) in each decile between 20 and 79 years, imaging them in a 1.5 Tesla CMR system. Scout acquisitions guided the placement of steady state free precession cine acquisitions transecting the main, right and left pulmonary arteries (MPA, RPA and LPA). Cross sections were rarely quite circular. RESULTS: From all subjects, the means of the greater and lesser orthogonal diastolic diameters in mm were: MPA, 22.9 ± 2.4 (M) and 21.2 ± 2.1 (F), RPA 16.6 ± 2.8 (M) and 14.7 ± 2.2 (F), and LPA 17.3 ± 2.5 (M) and 15.9 ± 2.0 (F), p < 0.0001 between genders in each case. The diastolic diameters increased with BSA and age, and plots are provided for reference. From measurements of minimum diastolic and maximum systolic cross sectional areas, the % systolic distensions were: MPA 42.7 ± 17.2 (M) and 41.8 ± 15.7 (F), RPA 50.6 ± 16.9 (M) and 48.2 ± 14.5 (F), LPA 35.6 ± 10.1 (M) and 35.2 ± 10.3 (F), and there was a decrease in distension with age (p < 0.0001 for the MPA). CONCLUSIONS: Measurements of MPA, RPA and LPA by cine CMR are provided for reference, with documentation of their changes with age and BSA.

Feasibility of cardiovascular magnetic resonance derived coronary wave intensity analysis.

BACKGROUND: Wave intensity analysis (WIA) of the coronary arteries allows description of the predominant mechanisms influencing coronary flow over the cardiac cycle. The data are traditionally derived from pressure and velocity changes measured invasively in the coronary artery. Cardiovascular magnetic resonance (CMR) allows measurement of coronary velocities using phase velocity mapping and derivation of central aortic pressure from aortic distension. We assessed the feasibility of WIA of the coronary arteries using CMR and compared this to invasive data. METHODS: CMR scans were undertaken in a serial cohort of patients who had undergone invasive WIA. Velocity maps were acquired in the proximal left anterior descending and proximal right coronary artery using a retrospectively-gated breath-hold spiral phase velocity mapping sequence with high temporal resolution (19 ms). A breath-hold segmented gradient echo sequence was used to acquire through-plane cross sectional area changes in the proximal ascending aorta which were used as a surrogate of an aortic pressure waveform after calibration with brachial blood pressure measured with a sphygmomanometer. CMR-derived aortic pressures and CMR-measured velocities were used to derive wave intensity. The CMR-derived wave intensities were compared to invasive data in 12 coronary arteries (8 left, 4 right). Waves were presented as absolute values and as a % of total wave intensity. Intra-study reproducibility of invasive and non-invasive WIA was assessed using Bland-Altman analysis and the intraclass correlation coefficient (ICC). RESULTS: The combination of the CMR-derived pressure and velocity data produced the expected pattern of forward and backward compression and expansion waves. The intra-study reproducibility of the CMR derived wave intensities as a % of the total wave intensity (mean ± standard deviation of differences) was 0.0 ± 6.8%, ICC = 0.91. Intra-study reproducibility for the corresponding invasive data was 0.0 ± 4.4%, ICC = 0.96. The invasive and CMR studies showed reasonable correlation (r = 0.73) with a mean difference of 0.0 ± 11.5%. CONCLUSION: This proof of concept study demonstrated that CMR may be used to perform coronary WIA non-invasively with reasonable reproducibility compared to invasive WIA. The technique potentially allows WIA to be performed in a wider range of patients and pathologies than those who can be studied invasively.

Dynamic inversion time for improved 3D late gadolinium enhancement imaging in patients with atrial fibrillation.

PURPOSE: High resolution three-dimensional (3D) late gadolinium enhancement (LGE) imaging is performed with single R-wave gating to minimize lengthy acquisition durations. In patients with atrial fibrillation (AF), heart rate variability results in variable magnetization recovery between sequence repeats, and image quality is often poor. In this study, we implemented and tested a dynamic inversion time (dynamic-TI) scheme designed to reduce sequence sensitivity to heart rate variations. METHODS: An inversion-prepared 3D segmented gradient echo sequence was modified so that the TI varied automatically from beat-to-beat (dynamic-TI) based on the time since the last sequence repeat. 3D LGE acquisitions were performed in 17 patients prior to radio frequency ablation of persistent AF both with and without dynamic-TI. Qualitative image quality scores, blood signal-to-ghosting ratios (SGRs). and blood-myocardium contrast-to-ghosting ratios (CGRs) were compared. RESULTS: Image quality scores were higher with dynamic-TI than without dynamic-TI (2.2 ± 0.9 vs. 1.8 ± 1.1, P = 0.008), as were blood-myocardium CGRs (13.8 ± 7.6 vs. 8.3 ± 6.1, P = 0.003) and blood SGRs (19.6 ± 8.5 vs. 13.1 ± 8.0, P = 0.003). CONCLUSION: The dynamic-TI algorithm improves image quality of 3D LGE imaging in this difficult patient population by reducing the sequence sensitivity to RR interval variations

Coronary artery wall imaging.

Like X-Ray contrast angiography, MR coronary angiograms show the vessel lumens rather than the vessels themselves. Consequently, outward remodeling of the vessel wall, which occurs in subclinical coronary disease before luminal narrowing, cannot be seen. The current gold standard for assessing the coronary vessel wall is intravascular ultrasound, and more recently, optical coherence tomography, both of which are invasive and use ionizing radiation. A noninvasive, low-risk technique for assessing the vessel wall would be beneficial to cardiologists interested in the early detection of preclinical disease and for the safe monitoring of the progression or regression of disease in longitudinal studies. In this review article, the current state of the art in MR coronary vessel wall imaging is discussed, together with validation studies and recent developments.

Free-breathing 3D late gadolinium enhancement imaging of the left ventricle using a stack of spirals at 3T.

PURPOSE: To develop navigator-gated free-breathing 3D spiral late gadolinium enhancement (LGE) imaging of the left ventricle at 3T and compare it with conventional breath-hold 2D Cartesian imaging. MATERIALS AND METHODS: Equivalent slices from 3D spiral and multislice 2D Cartesian acquisitions were compared in 15 subjects in terms of image quality (1, nondiagnostic to 5, excellent), sharpness (1-3), and presence of artifacts (0-2). Blood signal-to-noise ratio (SNR), blood/myocardium contrast-to-noise ratio (CNR), and quantitative sharpness were also compared. RESULTS: All 3D spiral scans were completed faster than an equivalent 2D Cartesian short-axis stack (85 vs. 230 sec, P < 0.001). Image quality was significantly higher for 2D Cartesian images than 3D spiral images (3.7 ± 0.87 vs. 3.4 ± 1.05, P = 0.03) but not for mid or apical slices specifically. There were no significant differences in qualitative and quantitative sharpness (95% confidence interval [CI]: 1.91 ± 0.67 vs. 1.93 ± 0.69, P = 0.83 and 95% CI: 0.41 ± 0.07 vs. 0.40 ± 0.09, P = 0.25, respectively), artifact scores (95% CI: 0.16 ± 0.37 vs. 0.40 ± 0.58, P = 0.16), SNR (95% CI: 121.5 ± 55.3 vs. 136.4 ± 77.9, P = 0.13), and CNR (95% CI: 101.6 ± 48.4 vs. 102.7 ± 61.8, P = 0.98). Similar enhancement ratios (0.65 vs. 0.62) and volumes (13.8 vs. 14.1cm(3) ) were measured from scar regions of three patients. CONCLUSIO: Navigator-gated 3D spiral LGE imaging can be performed in significantly and substantially shorter acquisition durations, although with some reduced image quality, than multiple breath-hold 2D Cartesian imaging while providing higher resolution and contiguous coverage.

Inter-study reproducibility of interleaved spiral phase velocity mapping of renal artery haemodynamics.

BACKGROUND: Qualitative and quantitative assessment of renal blood flow is valuable in the evaluation of patients with renal and renovascular diseases as well as in patients with heart failure. The temporal pattern of renal flow velocity through the cardiac cycle provides important information about renal haemodynamics. High temporal resolution interleaved spiral phase velocity mapping could potentially be used to study temporal patterns of flow and measure resistive and pulsatility indices which are measures of downstream resistance. METHODS: A retrospectively gated breath-hold spiral phase velocity mapping sequence (TR 19 ms) was developed at 3 Tesla. Phase velocity maps were acquired in the proximal right and left arteries of 10 healthy subjects in each of two separate scanning sessions. Each acquisition was analysed by two independent observers who calculated the resistive index (RI), the pulsatility index (PI), the mean flow velocity and the renal artery blood flow (RABF). Inter-study and inter-observer reproducibility of each variable was determined as the mean +/- standard deviation of the differences between paired values. The effect of background phase errors on each parameter was investigated. RESULTS: RI, PI, mean velocity and RABF per kidney were 0.71+/- 0.06, 1.47 +/- 0.29, 253.5 +/- 65.2 mm/s and 413 +/- 122 ml/min respectively. The inter-study reproducibilities were: RI -0.00 +/- 0.04 , PI -0.03 +/- 0.17, mean velocity -6.7 +/- 31.1 mm/s and RABF per kidney 17.9 +/- 44.8 ml/min. The effect of background phase errors was negligible (<2% for each parameter). CONCLUSIONS: High temporal resolution breath-hold spiral phase velocity mapping allows reproducible assessment of renal pulsatility indices and RABF.

Validation of high temporal resolution spiral phase velocity mapping of temporal patterns of left and right coronary artery blood flow against Doppler guidewire.

BACKGROUND: Temporal patterns of coronary blood flow velocity can provide important information on disease state and are currently assessed invasively using a Doppler guidewire. A non-invasive alternative would be beneficial as it would allow study of a wider patient population and serial scanning. METHODS: A retrospectively-gated breath-hold spiral phase velocity mapping sequence (TR 19 ms) was developed at 3 Tesla. Velocity maps were acquired in 8 proximal right and 15 proximal left coronary arteries of 18 subjects who had previously had a Doppler guidewire study at the time of coronary angiography. Cardiovascular magnetic resonance (CMR) velocity-time curves were processed semi-automatically and compared with corresponding invasive Doppler data. RESULTS: When corrected for differences in heart rate between the two studies, CMR mean velocity through the cardiac cycle, peak systolic velocity (PSV) and peak diastolic velocity (PDV) were approximately 40 % of the peak Doppler values with a moderate - good linear relationship between the two techniques (R(2): 0.57, 0.64 and 0.79 respectively). CMR values of PDV/PSV showed a strong linear relationship with Doppler values with a slope close to unity (0.89 and 0.90 for right and left arteries respectively). In individual vessels, plots of CMR velocities at all cardiac phases against corresponding Doppler velocities showed a consistent linear relationship between the two with high R(2) values (mean +/-SD: 0.79 +/-.13). CONCLUSIONS: High temporal resolution breath-hold spiral phase velocity mapping underestimates absolute values of coronary flow velocity but allows accurate assessment of the temporal patterns of blood flow.

Rapid automatic segmentation of abnormal tissue in late gadolinium enhancement cardiovascular magnetic resonance images for improved management of long-standing persistent atrial fibrillation.

BACKGROUND: Atrial fibrillation (AF) is the most common heart rhythm disorder. In order for late Gd enhancement cardiovascular magnetic resonance (LGE CMR) to ameliorate the AF management, the ready availability of the accurate enhancement segmentation is required. However, the computer-aided segmentation of enhancement in LGE CMR of AF is still an open question. Additionally, the number of centres that have reported successful application of LGE CMR to guide clinical AF strategies remains low, while the debate on LGE CMR's diagnostic ability for AF still holds. The aim of this study is to propose a method that reliably distinguishes enhanced (abnormal) from non-enhanced (healthy) tissue within the left atrial wall of (pre-ablation and 3 months post-ablation) LGE CMR data-sets from long-standing persistent AF patients studied at our centre. METHODS: Enhancement segmentation was achieved by employing thresholds benchmarked against the statistics of the whole left atrial blood-pool (LABP). The test-set cross-validation mechanism was applied to determine the input feature representation and algorithm that best predict enhancement threshold levels. RESULTS: Global normalized intensity threshold levels T PRE  = 1 1/4 and T POST  = 1 5/8 were found to segment enhancement in data-sets acquired pre-ablation and at 3 months post-ablation, respectively. The segmentation results were corroborated by using visual inspection of LGE CMR brightness levels and one endocardial bipolar voltage map. The measured extent of pre-ablation fibrosis fell within the normal range for the specific arrhythmia phenotype. 3D volume renderings of segmented post-ablation enhancement emulated the expected ablation lesion patterns. By comparing our technique with other related approaches that proposed different threshold levels (although they also relied on reference regions from within the LABP) for segmenting enhancement in LGE CMR data-sets of AF patients, we illustrated that the cut-off levels employed by other centres may not be usable for clinical studies performed in our centre. CONCLUSIONS: The proposed technique has great potential for successful employment in the AF management within our centre. It provides a highly desirable validation of the LGE CMR technique for AF studies. Inter-centre differences in the CMR acquisition protocol and image analysis strategy inevitably impede the selection of a universally optimal algorithm for segmentation of enhancement in AF studies.

Navigator artifact reduction in three-dimensional late gadolinium enhancement imaging of the atria.

PURPOSE: Navigator-gated three-dimensional (3D) late gadolinium enhancement (LGE) imaging demonstrates scarring following ablation of atrial fibrillation. An artifact originating from the slice-selective navigator-restore pulse is frequently present in the right pulmonary veins (PVs), obscuring the walls and making quantification of enhancement difficult. We describe a simple sequence modification to greatly reduce or remove this artifact. METHODS: A navigator-gated inversion-prepared gradient echo sequence was modified so that the slice-selective navigator-restore pulse was delayed in time from the nonselective preparation (NAV-restore-delayed). Both NAV-restore-delayed and conventional 3D LGE acquisitions were performed in 11 patients and the results compared. RESULTS: One patient was excluded due to severe respiratory motion artifact in both NAV-restore-delayed and conventional acquisitions. Moderate to severe artifact was present in 9 of the remaining 10 patients using the conventional sequence and was considerably reduced when using the NAV-restore-delayed sequence (ostial PV to blood pool ratio, 1.7 ± 0.5 versus 1.1 ± 0.2, respectively [P < 0.0001]; qualitative artifact scores, 2.8 ± 1.1 versus 1.2 ± 0.4, respectively [P < 0.001]). While navigator signal-to-noise ratio was reduced with the NAV-restore-delayed sequence, respiratory motion compensation was unaffected. CONCLUSIONS: Shifting the navigator-restore pulse significantly reduces or eliminates navigator artifact. This simple modification improves the quality of 3D LGE imaging and potentially aids late enhancement quantification in the atria.

Spiral tissue phase velocity mapping in a breath-hold with non-cartesian SENSE.

PURPOSE: Tissue phase velocity mapping (TPVM) is capable of reproducibly measuring regional myocardial velocities. However acquisition durations of navigator gated techniques are long and unpredictable while current breath-hold techniques have low temporal resolution. This study presents a spiral TPVM technique which acquires high resolution data within a clinically acceptable breath-hold duration. METHODS: Ten healthy volunteers are scanned using a spiral sequence with temporal resolution of 24 ms and spatial resolution of 1.7 × 1.7 mm. Retrospective cardiac gating is used to acquire data over the entire cardiac cycle. The acquisition is accelerated by factors of 2 and 3 by use of non-Cartesian SENSE implemented on the Gadgetron GPU system resulting in breath-holds of 17 and 13 heartbeats, respectively. Systolic, early diastolic, and atrial systolic global and regional longitudinal, circumferential, and radial velocities are determined. RESULTS: Global and regional velocities agree well with those previously reported. The two acceleration factors show no significant differences for any quantitative parameter and the results also closely match previously acquired higher spatial resolution navigator-gated data in the same subjects. CONCLUSION: By using spiral trajectories and non-Cartesian SENSE high resolution, TPVM data can be acquired within a clinically acceptable breath-hold.

Improved respiratory efficiency of 3D late gadolinium enhancement imaging using the continuously adaptive windowing strategy (CLAWS).

PURPOSE: Acquisition durations of navigator-gated high-resolution three-dimensional late gadolinium enhancement studies may typically be up to 10 min, depending on the respiratory efficiency and heart rate. Implementation of the continuously adaptive windowing strategy (CLAWS) could increase respiratory efficiency, but the resulting non-smooth k-space acquisition order during gadolinium wash-out could result in increased artifact. METHODS: Navigator-gated three-dimensional late gadolinium enhancement acquisitions were performed in 18 patients using tracking end-expiratory accept/reject (EE-ARA) and CLAWS algorithms in random order. RESULTS: Retrospective analysis of the stored navigator data shows that CLAWS scan times are very close to (within 1%) or equal to the fastest achievable scan times while EE-ARA significantly extends the acquisition duration (P < 0.0001). EE-ARA acquisitions are 26% longer than CLAWS acquisitions (378 ± 104 s compared to 301 ± 85 s, P = 0.002). Image quality scores for CLAWS and EE-ARA acquisitions are not significantly different (4.1 ± 0.6 compared to 4.3 ± 0.6, P = ns). Numerical phantom simulations show that the non-uniform k-space ordering introduced by CLAWS results in slight, but not statistically significant, reductions in both blood signal-to-noise ratio (10%) and blood-myocardium contrast-to-noise ratio (12%). CONCLUSIONS: CLAWS results in markedly reduced acquisition durations compared to EE-ARA without significant detriment to the image quality.

In vivo diffusion tensor MRI of the human heart: reproducibility of breath-hold and navigator-based approaches.

The aim of this study was to implement a quantitative in vivo cardiac diffusion tensor imaging (DTI) technique that was robust, reproducible, and feasible to perform in patients with cardiovascular disease. A stimulated-echo single-shot echo-planar imaging (EPI) sequence with zonal excitation and parallel imaging was implemented, together with a novel modification of the prospective navigator (NAV) technique combined with a biofeedback mechanism. Ten volunteers were scanned on two different days, each time with both multiple breath-hold (MBH) and NAV multislice protocols. Fractional anisotropy (FA), mean diffusivity (MD), and helix angle (HA) fiber maps were created. Comparison of initial and repeat scans showed good reproducibility for both MBH and NAV techniques for FA (P > 0.22), MD (P > 0.15), and HA (P > 0.28). Comparison of MBH and NAV FA (FAMBHday1 = 0.60 ± 0.04, FANAVday1 = 0.60 ± 0.03, P = 0.57) and MD (MDMBHday1 = 0.8 ± 0.2 × 10(-3) mm(2) /s, MDNAVday1 = 0.9 ± 0.2 × 10(-3) mm(2) /s, P = 0.07) values showed no significant differences, while HA values (HAMBHday1Endo = 22 ± 10°, HAMBHday1Mid-Endo = 20 ± 6°, HAMBHday1Mid-Epi = -1 ± 6°, HAMBHday1Epi = -17 ± 6°, HANAVday1Endo = 7 ± 7°, HANAVday1Mid-Endo = 13 ± 8°, HANAVday1Mid-Epi = -2 ± 7°, HANAVday1Epi = -14 ± 6°) were significantly different. The scan duration was 20% longer with the NAV approach. Currently, the MBH approach is the more robust in normal volunteers. While the NAV technique still requires resolution of some bulk motion sensitivity issues, these preliminary experiments show its potential for in vivo clinical cardiac diffusion tensor imaging and for delivering high-resolution in vivo 3D DTI tractography of the heart.

MR assessment of regional myocardial mechanics.

Regional myocardial function can be measured by several MR techniques including tissue tagging, phase velocity mapping, and more recently, displacement encoding with stimulated echoes (DENSE) and strain encoding (SENC). Each of these techniques was developed separately and has undergone significant change since its original implementation. As a result, in the current literature, the common features and the differences between the techniques and what they measure are often unclear and confusing. This review article delivers an extensively referenced introductory text which clarifies the current methodology from the starting point of the Bloch equations. By doing this in a consistent way for each method, the similarities and differences between them are highlighted. In addition, their capabilities and limitations are discussed, together with their relative advantages and disadvantages. While the focus is on sequence design and development, the principal parameters measured by each technique are also summarized, together with brief results, with the reader being directed to the extensive literature on data processing and clinical applications for more detail.

Efficient and reproducible high resolution spiral myocardial phase velocity mapping of the entire cardiac cycle.

BACKGROUND: Three-directional phase velocity mapping (PVM) is capable of measuring longitudinal, radial and circumferential regional myocardial velocities. Current techniques use Cartesian k-space coverage and navigator-gated high spatial and high temporal resolution acquisitions are long. In addition, prospective ECG-gating means that analysis of the full cardiac cycle is not possible. The aim of this study is to develop a high temporal and high spatial resolution PVM technique using efficient spiral k-space coverage and retrospective ECG-gating. Detailed analysis of regional motion over the entire cardiac cycle, including atrial systole for the first time using MR, is presented in 10 healthy volunteers together with a comprehensive assessment of reproducibility. METHODS: A navigator-gated high temporal (21 ms) and spatial (1.4 × 1.4 mm) resolution spiral PVM sequence was developed, acquiring three-directional velocities in 53 heartbeats (100% respiratory-gating efficiency). Basal, mid and apical short-axis slices were acquired in 10 healthy volunteers on two occasions. Regional and transmural early systolic, early diastolic and atrial systolic peak longitudinal, radial and circumferential velocities were measured, together with the times to those peaks (TTPs). Reproducibilities were determined as mean ± SD of the signed differences between measurements made from acquisitions performed on the two days. RESULTS: All slices were acquired in all volunteers on both occasions with good image quality. The high temporal resolution allowed consistent detection of fine features of motion, while the high spatial resolution allowed the detection of statistically significant regional and transmural differences in motion. Colour plots showing the regional variations in velocity over the entire cardiac cycle enable rapid interpretation of the regional motion within any given slice. The reproducibility of peak velocities was high with the reproducibility of early systolic, early diastolic and atrial systolic peak radial velocities in the mid slice (for example) being -0.01 ± 0.36, 0.20 ± 0.56 and 0.14 ± 0.42 cm/s respectively. Reproducibility of the corresponding TTP values, when normalised to a fixed systolic and diastolic length, was also high (-13.8 ± 27.4, 1.3 ± 21.3 and 3.0 ± 10.9 ms for early systolic, early diastolic and atrial systolic respectively). CONCLUSIONS: Retrospectively gated spiral PVM is an efficient and reproducible method of acquiring 3-directional, high resolution velocity data throughout the entire cardiac cycle, including atrial systole.

Fully-modelled blood-focused variable inversion times for 3D late gadolinium-enhanced imaging.

PURPOSE: Variable heart rate during single-cycle inversion-recovery Late Gadolinium-Enhanced (LGE) scanning degrades image quality, which can be mitigated using Variable Inversion Times (VTIs) in real-time response to R-R interval changes. We investigate in vivo and in simulations an extension of a single-cycle VTI method previously applied in 3D LGE imaging, that now fully models the longitudinal magnetisation (fmVTI). METHODS: The VTI and fmVTI methods were used to perform 3D LGE scans for 28 3D LGE patients, with qualitative image quality scores assigned for left atrial wall clarity and total ghosting. Accompanying simulations of numerical phantom images were assessed in terms of ghosting of normal myocardium, blood, and myocardial scar. RESULTS: The numerical simulations for fmVTI showed a significant decrease in blood ghosting (VTI: 410 ± 710, fmVTI: 68 ± 40, p < 0.0005) and scar ghosting (VTI: 830 ± 1300, fmVTI: 510 ± 730, p < 0.02). Despite this, there was no significant change in qualitative image quality scores, either for left atrial wall clarity (VTI: 2.0 ± 1.0, fmVTI: 1.8 ± 1.0, p > 0.1) or for total ghosting (VTI: 1.9 ± 1.0, fmVTI: 2.0 ± 1.0, p > 0.7). CONCLUSIONS: Simulations indicated reduced ghosting with the fmVTI method, due to reduced Mz variability in the blood signal. However, other sources of phase-encode ghosting and blurring appeared to dominate and obscure this finding in the patient studies available.

Role of cardiovascular magnetic resonance as a gatekeeper to invasive coronary angiography in patients presenting with heart failure of unknown etiology.

BACKGROUND: In patients presenting with new-onset heart failure of uncertain etiology, the role of coronary angiography (CA) is unclear. Although conventionally performed to differentiate underlying coronary artery disease from dilated cardiomyopathy, CA is associated with a risk of complications and may not detect an ischemic cause resulting from arterial recanalization or an embolic episode. In this study, we assessed the diagnostic accuracy of a cardiovascular magnetic resonance (CMR) protocol incorporating late gadolinium enhancement (LGE) and magnetic resonance CA as a noninvasive gatekeeper to CA in determining the etiology of heart failure in this subset of patients. METHODS AND RESULTS: One hundred twenty consecutive patients underwent CMR and CA. The etiology was ascribed by a consensus panel that used the results of the CMR scans. Similarly, a separate consensus group ascribed an underlying cause by using the results of CA. The diagnostic accuracy of both strategies was compared against a gold-standard panel that made a definitive judgment by reviewing all clinical data. The study was powered to show noninferiority between the 2 techniques. The sensitivity of 100%, specificity of 96%, and diagnostic accuracy of 97% for LGE-CMR were equivalent to CA (sensitivity, 93%; specificity, 96%; and diagnostic accuracy, 95%). As a gatekeeper to CA, LGE-CMR was also found to be a cheaper diagnostic strategy in a decision tree model when United Kingdom-based costs were assumed. The economic merits of this model would change, depending on the relative costs of LGE-CMR and CA in any specific healthcare system. CONCLUSION: This study showed that LGE-CMR is a safe, clinically effective, and potentially economical gatekeeper to CA in patients presenting with heart failure of uncertain etiology.

Adaptive Hierarchical Dual Consistency for Semi-Supervised Left Atrium Segmentation on Cross-Domain Data.

Semi-supervised learning provides great significance in left atrium (LA) segmentation model learning with insufficient labelled data. Generalising semi-supervised learning to cross-domain data is of high importance to further improve model robustness. However, the widely existing distribution difference and sample mismatch between different data domains hinder the generalisation of semi-supervised learning. In this study, we alleviate these problems by proposing an Adaptive Hierarchical Dual Consistency (AHDC) for the semi-supervised LA segmentation on cross-domain data. The AHDC mainly consists of a Bidirectional Adversarial Inference module (BAI) and a Hierarchical Dual Consistency learning module (HDC). The BAI overcomes the difference of distributions and the sample mismatch between two different domains. It mainly learns two mapping networks adversarially to obtain two matched domains through mutual adaptation. The HDC investigates a hierarchical dual learning paradigm for cross-domain semi-supervised segmentation based on the obtained matched domains. It mainly builds two dual-modelling networks for mining the complementary information in both intra-domain and inter-domain. For the intra-domain learning, a consistency constraint is applied to the dual-modelling targets to exploit the complementary modelling information. For the inter-domain learning, a consistency constraint is applied to the LAs modelled by two dual-modelling networks to exploit the complementary knowledge among different data domains. We demonstrated the performance of our proposed AHDC on four 3D late gadolinium enhancement cardiac MR (LGE-CMR) datasets from different centres and a 3D CT dataset. Compared to other state-of-the-art methods, our proposed AHDC achieved higher segmentation accuracy, which indicated its capability in the cross-domain semi-supervised LA segmentation.