Regional myocardial strain by cardiac magnetic resonance feature tracking for detection of scar in ischemic heart disease

BackgroundAlthough cardiac magnetic resonance (CMR) can accurately quantify global left ventricular strain using feature tracking (FT), it has been suggested that FT cannot reliably quantify regional strain. We aimed to determine whether abnormalities in regional strain measured using FT can be detected within areas of myocardial scar and to determine the extent to which the regional strain measurement is impacted by LV ejection fraction (EF).MethodsWe retrospectively studied 96 patients (46 with LVEF ≤ 40%, 50 with LVEF > 40%) with coronary artery disease and a late gadolinium enhancement (LGE) pattern consistent with myocardial infarction, who underwent CMR imaging (1.5T). Regional peak systolic longitudinal and circumferential strains (RLS, RCS) were measured within LGE and non-LGE areas. Linear regression analysis was performed for strain in both areas against LVEF to determine whether the relationship between strain and LGE holds across the LV function spectrum. Receiver-operating curve (ROC) analysis was performed in 33 patients (derivation cohort) to optimize strain cutoff, which was tested in the remaining 63 patients (validation cohort) for its ability to differentiate LGE from non-LGE areas.ResultsBoth RLS and RCS magnitudes were reduced in LGE areas: RLS = −10.4 ± 6.2% versus −21.0 ± 8.5% (p < 0.001); RCS = −10.4 ± 6.0% versus −18.9 ± 8.6%, respectively (p < 0.001), but there was considerable overlap between LGE and non-LGE areas. Linear regression revealed that it was partially driven by the natural dependence between strain and EF, suggesting that EF-corrected strain cutoff is needed to detect LGE. ROC analysis showed the ability of both RLS and RCS to differentiate LGE from non-LGE areas: area under curve 0.95 and 0.89, respectively. In the validation cohort, optimal cutoffs of RLS/EF = 0.36 and RCS/EF = 0.37 yielded sensitivity, specificity and accuracy 0.74–0.78.ConclusionAbnormalities in RLS and RCS within areas of myocardial scar can be detected using CMR-FT; however, LVEF must be accounted for.     

Evaluation of left ventricular function with cardiac magnetic resonance imaging using Fourier fitting

The aims of this study were to investigate the applicability of Fourier fitting in the magnetic resonance (MR) evaluation of left ventricular (LV) function and to determine the optimal number of harmonics for fitting. Cine cardiac MR imaging was performed in 10 subjects, and an LV time–volume curve was generated. Fourier fitting was applied to the original curve using 1–10 harmonics, and the qualities of the time–volume curve and first-derivative curve were evaluated. LV functional parameters were calculated from curves generated with and without fitting. The quality of the original time–volume curve was good, and Fourier fitting had no substantial effect on functional parameters obtained directly from the time–volume curve such as ejection fraction. The first-derivative curve generated without fitting showed substantial artificial fluctuation. The application of Fourier fitting depressed the fluctuation and tended to decrease estimates of peak ejection rate and peak filling rate. Five or six harmonics appeared to be appropriate for obtaining a high-quality first-derivative curve. In conclusion, Fourier fitting was indicated to aid in reducing the artificial fluctuation of the first-derivative curve generated from cine cardiac MR imaging and to contribute to the evaluation of functional parameters derived from the first-derivative curve.     

A deep-learning semantic segmentation approach to fully automated MRI-based left-ventricular deformation analysis in cardiotoxicity

Left-ventricular (LV) strain measurements with the Displacement Encoding with Stimulated Echoes (DENSE) MRI sequence provide accurate estimates of cardiotoxicity damage related to breast cancer chemotherapy. This study investigated an automated LV chamber quantification tool via segmentation with a supervised deep convolutional neural network (DCNN) before strain analysis with DENSE images. Segmentation for chamber quantification analysis was conducted with a custom DeepLabV3+ DCNN with ResNet-50 backbone on 42 female breast cancer datasets (22 training-sets, eight validation-sets and 12 independent test-sets). Parameters such as LV end-diastolic diameter (LVEDD) and ejection fraction (LVEF) were quantified, and myocardial strains analyzed with the Radial Point Interpolation Method (RPIM). Myocardial classification was validated against ground-truth with sensitivity-specificity analysis, the metrics of Dice, average perpendicular distance (APD) and Hausdorff-distance. Following segmentation, validation was conducted with the Cronbach's Alpha (C-Alpha) intraclass correlation coefficient between LV chamber quantification results with DENSE and Steady State Free Precession (SSFP) acquisitions and a vendor tool-based method to segment the DENSE data, and similarly for myocardial strain analysis in the chambers. The results of myocardial classification from segmentation of the DENSE data were accuracy = 97%, Dice = 0.89 and APD = 2.4 mm in the test-set. The C-Alpha correlations from comparing chamber quantification results between the segmented DENSE and SSFP data and vendor tool-based method were 0.97 for LVEF (56 ± 7% vs 55 ± 7% vs 55 ± 6%, p = 0.6) and 0.77 for LVEDD (4.6 ± 0.4 cm vs 4.5 ± 0.3 cm vs 4.5 ± 0.3 cm, p = 0.8). The validation metrics against ground-truth and equivalent parameters obtained from the SSFP segmentation and vendor tool-based comparisons show that the DCNN approach is applicable for automated LV chamber quantification and subsequent strain analysis in cardiotoxicity.     

Left ventricular volume measurements with free breathing respiratory self-gated 3-dimensional golden angle radial whole-heart cine imaging – Feasibility and reproducibility

PurposeTo develop and evaluate a free breathing respiratory self-gated isotropic resolution technique for left ventricular (LV) volume measurements.MethodsA 3D radial trajectory with double golden-angle ordering was used for free-running data acquisition during free breathing in 9 healthy volunteers. A respiratory self-gating signal was extracted from the center of k-space and used with the electrocardiogram to bin all data into 3 respiratory and 25 cardiac phases. 3D image volumes were reconstructed and the LV endocardial border was segmented. LV volume measurements and reproducibility from 3D free breathing cine were compared to conventional 2D breath-held cine.ResultsNo difference was found between 3D free breathing cine and 2D breath-held cine with regards to LV ejection fraction, stroke volume, end-systolic volume and end-diastolic volume (P < 0.05 for all). The test-retest differences did not differ between 3D free breathing cine and 2D breath-held cine (P < 0.05 for all).Conclusion3D free breathing cine and conventional 2D breath-held cine showed similar values and test-retest repeatability for LV volumes in healthy volunteers. 3D free breathing cine enabled retrospective sorting and arbitrary angulation of isotropic data, and could correctly measure LV volumes during free breathing acquisition.     

Tissue-tracking in the assessment of late gadolinium enhancement in myocarditis and myocardial infarction

PurposeTo test the diagnostic performance of cardiovascular magnetic resonance (CMR) tissue-tracking (TT) to detect the presence of late gadolinium enhancement (LGE) in patients with a diagnosis of myocardial infarction (MI) or myocarditis (MYO), preserved left ventricular ejection fraction (LVEF) and no visual regional wall motion abnormalities (RWMA).MethodsWe selected consecutive CMR studies of 50 MI, 50 MYO and 96 controls. Receiving operating characteristic (ROC) curve and net reclassification index (NRI) analyses were used to assess the predictive ability and the incremental diagnostic yield of 2D and 3D TT-derived strain parameters for the detection of LGE and to measure the best cut-off values of strain parameters.ResultsOverall, cases showed significantly reduced 2D global longitudinal strain (2D-GLS) values compared with controls (−20.1 ± 3.1% vs −21.6 ± 2.7%; p = 0.0008). 2D-GLS was also significantly reduced in MYO patients compared with healthy controls (−19.7 ± 2.9% vs −21.9 ± 2.4%; p = 0.0001). 3D global radial strain (3D-GRS) was significantly reduced in MI patients compared with controls with risk factors (34.3 ± 11.8% vs 40.3 ± 12.5%, p = 0.024) Overall, 2D-GLS yielded good diagnostic accuracy for the detection of LGE in the MYO subgroup (AUROC 0.79; NRI (95% CI) = 0.6 (0.3, 1.02) p = 0.0004), with incremental predictive value beyond risk factors and LV function parameters (p for AUROC difference = 0.048). In the MI subgroup, 2D-GRS (AUROC 0.81; NRI (95% CI) = 0.56 (0.17, 0.95) p = 0.004), 3D-GRS (AUROC 0.82; NRI (95% CI) = 0.57 (0.17, 0.97) p = 0.006) and 3D global circumferential strain (3D-GCS) (AUROC 0.81; NRI (95% CI) = 0.62 (0.22, 1.01) p = 0.002) emerged as potential markers of disease. The best cut-off for 2D-GLS was −21.1%, for 2D- and 3D-GRS were 39.1% and 37.7%, respectively, and for 3D-GCS was −16.4%.ConclusionsAt CMR-tissue tracking analysis, 2D-GLS was a significant predictor of LGE in patients with myocarditis but preserved LVEF and no visual RWMA. Both 2D- and 3D-GRS and 2D-GCS yielded good diagnostic accuracy for LGE detection in patients with previous MI but preserved LVEF and no visual RWMA.     

A simplified method for the determination of left atrial size and function using cine magnetic resonance imaging

This work aimed at developing a rapid and clinically applicable method for the assessment of left atrial size and function using magnetic resonance imaging (MRI). We studied 17 healthy subjects and 26 cardiac patients. Left atrial cine MRI with 50 ms phases was made in 6–12 contiguous long-axis sections encompassing the entire atrial cavity. A volume-time curve was reconstructed to measure the minimum and maximum volumes as well as the fractional volume change, reservoir function, ejection fraction, and mean filling and emptying rates of the left atrium. The image section with the largest left atrial area was then selected and a comparable area-time curve was reconstructed. The atrial phasic areas and functional indices were determined analogously to the volume-based assessment. The contours of atrial area-time and volume-time curves agreed closely in individual subjects. All area-based left atrial measurements distinguished cardiac patients as a group from healthy persons. The combined specificity of the area-based analyses was 92% and the sensitivity, 65%, in identifying abnormal results in individual patients. The accuracy of the area-based data was best for the atrial minimum size, fractional change, reservoir function, and mean filling rate. The estimated time savings with the simplified method were 5 to 6 h per patient. Left atrial size and function can be studied by reconstructing a phasic atrial area-time curve with cine MRI. Atrial enlargement and abnormalities of filling and reservoir function can be reliably identified, but if data on conduit or stroke function are crucial the three-dimensional MRI technique is still recommended.     

Non-contrast enhanced diagnosis of acute myocarditis based on the 17-segment heart model using 2D-feature tracking magnetic resonance imaging

PurposeThe aim of this study was to investigate the diagnostic value of myocardial deformation analysis based on the 17-segment heart model using non-contrast enhanced (CE) 2D tissue feature tracking (2D-FT) technique.Material and methodsSeventy patients with suspected myocarditis underwent a cardiovascular magnetic resonance (CMR) examination at 1.5 Tesla. A contrast-agent-free part of this CMR protocol was additionally performed in forty healthy volunteers (HV). Besides standard CMR data sets, 2D-FT derived segmental and global longitudinal, radial, and circumferential deformation parameters were analyzed. The 2D-FT results were compared to the combined findings from CMR imaging and endomyocardial biopsy (EMB).ResultsPatients were assigned to three groups depending on their ejection fraction (EF) (<40%, 40–55%, ≥55%). Compared to HV, impaired EF (<55%) was significantly correlated to reduced segmental and global strain and strain rate values. The circumferential deformation analysis was more sensitive to myocardial changes than longitudinal and radial analysis. The segmental strain/strain rate had an accuracy of 84.3%/70.0% for the diagnosis of an acute myocarditis, stated by EMB and CMR in 42 of 70 patients. In patients with preserved EF, acute myocarditis could be ruled out using only segmental strain analysis with a negative predictive value of 87.5%.ConclusionIn patients with suspected myocarditis, the deformation analysis based on the 17-segment heart model provides valuable information about functional myocardial inhomogeneity. This quantitative approach could be used in addition to the clinical standard CMR protocol and represents a promising tool in the framework of a prospective automatized multiparametric CMR imaging analysis.     

Automated measurement of fat infiltration in the hip abductors from Dixon magnetic resonance imaging

PurposeIntramuscular fat infiltration is a dynamic process, in response to exercise and muscle health, which can be quantified by estimating fat fraction (FF) from Dixon MRI. Healthy hip abductor muscles are a good indicator of a healthy hip and an active lifestyle as they have a fundamental role in walking. The automated measurement of the abductors' FF requires the challenging task of segmenting them. We aimed to design, develop and evaluate a multi-atlas based method for automated measurement of fat fraction in the main hip abductor muscles: gluteus maximus (GMAX), gluteus medius (GMED), gluteus minimus (GMIN) and tensor fasciae latae (TFL).MethodWe collected and manually segmented Dixon MR images of 10 healthy individuals and 7 patients who underwent MRI for hip problems. Twelve of them were selected to build an atlas library used to implement the automated multi-atlas segmentation method. We compared the FF in the hip abductor muscles for the automated and manual segmentations for both healthy and patients groups. Measures of average and spread were reported for FF for both methods. We used the root mean square error (RMSE) to quantify the method accuracy. A linear regression model was used to explain the relationship between FF for automated and manual segmentations.ResultsThe automated median (IQR) FF was 20.0(16.0–26.4) %, 14.3(10.9–16.5) %, 15.5(13.9–18.6) % and 16.2(13.5–25.6) % for GMAX, GMED, GMIN and TFL respectively, with a FF RMSE of 1.6%, 0.8%, 2.1%, 2.7%. A strong linear correlation (R² = 0.93, p < .001, m = 0.99) was found between the FF from automated and manual segmentations. The mean FF was higher in patients than in healthy subjects.ConclusionThe automated measurement of FF of hip abductor muscles from Dixon MRI had good agreement with FF measurements from manually segmented images. The method was accurate for both healthy and patients groups.     

A non-contrast CMR index for assessing myocardial fibrosis

PurposeSafe, sensitive, and non-invasive imaging methods to assess the presence, extent, and turnover of myocardial fibrosis are needed for early stratification of risk in patients who might develop heart failure after myocardial infarction. We describe a non-contrast cardiac magnetic resonance (CMR) approach for sensitive detection of myocardial fibrosis using a canine model of myocardial infarction and reperfusion.MethodsSeven dogs had coronary thrombotic occlusion of the left anterior descending coronary arteries followed by fibrinolytic reperfusion. CMR studies were performed at 7 days after reperfusion. A CMR spin-locking T₁ρ mapping sequence was used to acquire T₁ρ dispersion data with spin-lock frequencies of 0 and 511 Hz. A fibrosis index map was derived on a pixel-by-pixel basis. CMR native T₁ mapping, first-pass myocardial perfusion imaging, and post-contrast late gadolinium enhancement imaging were also performed for assessing myocardial ischemia and fibrosis. Hearts were dissected after CMR for histopathological staining and two myocardial tissue segments from the septal regions of adjacent left ventricular slices were qualitatively assessed to grade the extent of myocardial fibrosis.ResultsHistopathology of 14 myocardial tissue segments from septal regions was graded as grade 1 (fibrosis area, < 20% of a low power field, n = 9), grade 2 (fibrosis area, 20–50% of field, n = 4), or grade 3 (fibrosis area, > 50% of field, n = 1). A dramatic difference in fibrosis index (183%, P < 0.001) was observed by CMR from grade 1 to 2, whereas differences were much smaller for T₁ρ (9%, P = 0.14), native T₁ (5.5%, P = 0.12), and perfusion (− 21%, P = 0.05).ConclusionA non-contrast CMR index based on T₁ρ dispersion contrast was shown in preliminary studies to detect and correlate with the extent of myocardial fibrosis identified histopathologically. A non-contrast approach may have important implications for managing cardiac patients with heart failure, particularly in the presence of impaired renal function.     

Single-shot morpho-functional and structural characterization of the left-ventricle in a mouse model of acute ischemia-reperfusion injury with an optimized 3D IntraGate cine FLASH sequence at 7T MR

Preclinical cardiac MR is challenging and time-consuming. A fast and comprehensive acquisition protocol and standardized image post-processing may improve preclinical research, reducing acquisition time, costs and variability of results. In the present study, we evaluated the feasibility of a contrast-enhanced 3D IntraGate steady-state cine sequence (ce-3D-IG-cine) with short acquisition time (11 min) for a single-shot combined characterization of left ventricle (LV) remodeling and infarct size (IS) in a mouse model of acute ischemia-reperfusion injury.Sixteen male C57BL/6N mice underwent 7T cardiac MR (Bruker, BioSpec 70/30) including optimized ce-3D-IG-cine (total scan time 11 min) at day 1, 5 and 28 after surgery. LV end-diastolic volume (EDV_MR) and ejection fraction (EF_MR) extracted from MR were compared to ones from short-axis (SA-EDV_echo, SA-EF_echo) and parasternal long-axis (LA-EDV_echo, LA-EF_echo) echocardiography. IS was manually and semiautomatically segmented from ce-3D-IG-cine using different standard deviation (SD +2, +3, +4, +5, +6 in respect to a reference tissue).Mice were sacrificed at day 28, immediately after imaging. IS at day 28 was compared to injury burden at histology.MR and echocardiographic morpho-functional parameters were compared, as IS from MR and histology. Bland-Altman plots were used to assess the agreement in ischemic burden segmentation.Volumetric and functional parameters measured on ce-3D-IG-cine correlated to the correspondent echocardiographic parameter (EDV_MR vs SA-EDV_echo: ρ = 0.813; EDV_MR vs LA-EDV_echo: ρ = 0.845; EF_MR vs SA-EF_echo ρ = 0.612; EF_MR vs LA-EF_echo ρ = 0.791; p < 0.001 in all cases).Manually segmented IS strongly correlated with the scar at histology (ρ = 0.904, p < 0.001). A threshold of +3SD showed the highest performance for semiautomatic assessment of IS compared to manual segmentation (ρ = 0.965, p < 0.001), with an overall reproducibility of 73%, and a peak reproducibility of 80% at day 1.The ce-3D-IG-cine sequence, manually or semiautomatically segmented using 3SD threshold, allows fast and comprehensive LV morpho-functional and structural characterization in myocardial ischemia-reperfusion injury model.     

Mapping of left ventricle wall thickness in mice using 11.7-T magnetic resonance imaging

BackgroundThe left ventricle (LV) wall thickness is an important and routinely measured cardiologic parameter. Here we introduce three-dimensional (3D) mapping of LV wall thickness and function using a self-gated magnetic resonance (MR) sequence for ultra-high-field 11.7-T MR cine imaging of mouse hearts.Methods and resultsSix male C57BL/6-j mice were subjected to 11.7-T MR imaging (MRI). Three standard views—short axis, long axis four-chamber, and long axis two-chamber—and eight consecutive short axis scans from the apex to base were performed for each mouse. The resulting 11 self-gated cine images were used for fast low-angle shot analysis with a navigator echo over an observation period of approximately 35 min. The right ventricle (RV) and LV were identified in the short axis and four-chamber views. On 3D color-coded maps, the interventricular septum wall (diastole: 0.94 ± 0.05 mm, systole: 1.20 ± 0.09 mm) and LV free wall (diastole: 1.07 ± 0.15 mm, systole: 1.79 ± 0.11 mm) thicknesses were measured.ConclusionThis 3D wall thickness mapping technique can be used to observe regional wall thickness at the end-diastole and end-systole. Self-gated cine imaging based on ultra-high-field MRI can be used to accurately and easily measure cardiac function and wall thickness in normal mouse hearts. As in the preclinical study, this versatile and simple method will be clinically useful for the high-field-MRI evaluation of cardiac function and wall thickness.     

Group delay method for MRI aortic pulse wave velocity measurements in clinical protocols with low temporal resolution: Validation in a heterogeneous cohort

BackgroundMRI assessment of aortic pulse wave velocity (PWV) helps predict the risk of vascular events, but the recommended phase contrast sampling rate is faster than what is utilized in most clinical sequences. There are many existing MRI databases obtained for assessment of cardiac output using lower temporal frequency sampling where information might be obtained about aortic stiffness (PWV). In this work, we sought to evaluate whether the Group Delay (GD) method can generate a reproducible measure of stiffness and describe expected age-related stiffening of the aortic arch using lower sampling rates in standard clinical sequences.MethodsPhase contrast (PC) MRI was obtained on the ascending and descending aortic arch in a heterogeneous adult cohort (n = 23; 9 women) spanning over a wide range of ages (ages 24–89, mean 49.4 ± 18.4). Data was collected with standard cardiac MRI protocols for cardiac output evaluation (repetition time = 7.8 ms, views-per-segment = 4, encoding velocity = 200 cm/s). Pulse wave transit times (TT) were computed using the GD method, two other validated automated approaches (cross correlation TT Algorithm by Gaddum and Segment by Medviso), and the manual tangent method. Pressure waveforms from tonometry and flow waveforms from PC MRI were used to assess wave reflections.ResultsGroup Delay and TT-Algorithm showed significant and high retest reproducibility (r = 0.86 for both) as well as high PWV correlation with age (r = 0.93, P-value < 0.00005 and r = 0.96, P-value < 0.00005 respectively) and with each other (r = 0.94, P-value < 0.00001, RMSE = 0.94 m/s). Arbitrary altering of the image acquisition trigger in the GD method introduced error of 10%–13%, but the TT-algorithm error range was 11%–25%.ConclusionGroup Delay enables reproducible assessment of transit time to derive PWV from low temporal resolution clinical cardiac MRI sequences that can also identify age-related stiffening.     

结合生物力学模型重建左心室位移场

左心室心肌最为发达,心肌收缩产生的高压将动脉血泵入全身,集中体现了心脏的泵血能力.定量分析左心室收缩运动是诊断心血管疾病（如心肌梗死）的重要途径.本文采用描述左心室心肌材质的生物力学模型重建左心室位移场.该力学模型作为插值项,与心脏电影磁共振图像的观测位移场共同纳入贝叶斯估计框架,并采用有限元法求解位移场方程.实验比较了左心室射血无力组（46例）与正常组（55例）的左心室功能参数,发现两组在径向和圆周方向的位移、速度、应变和应变率都具有非常显著的差异（p < 0.001）,这证明本文方法能够有效区别左心室运动正常与否.实验结果还与CVI软件测量的左心室功能参数具有较高的相关性,说明本文方法有望辅助心血管疾病的临床诊断.     

Accuracy and reproducibility of automated white matter hyperintensities segmentation with lesion segmentation tool: A European multi-site 3T study

Brain vascular damage accumulate in aging and often manifest as white matter hyperintensities (WMHs) on MRI. Despite increased interest in automated methods to segment WMHs, a gold standard has not been achieved and their longitudinal reproducibility has been poorly investigated. The aim of present work is to evaluate accuracy and reproducibility of two freely available segmentation algorithms. A harmonized MRI protocol was implemented in 3T-scanners across 13 European sites, each scanning five volunteers twice (test-retest) using 2D-FLAIR. Automated segmentation was performed using Lesion segmentation tool algorithms (LST): the Lesion growth algorithm (LGA) in SPM8 and 12 and the Lesion prediction algorithm (LPA). To assess reproducibility, we applied the LST longitudinal pipeline to the LGA and LPA outputs for both the test and retest scans. We evaluated volumetric and spatial accuracy comparing LGA and LPA with manual tracing, and for reproducibility the test versus retest. Median volume difference between automated WMH and manual segmentations (mL) was −0.22[IQR = 0.50] for LGA-SPM8, −0.12[0.57] for LGA-SPM12, −0.09[0.53] for LPA, while the spatial accuracy (Dice Coefficient) was 0.29[0.31], 0.33[0.26] and 0.41[0.23], respectively. The reproducibility analysis showed a median reproducibility error of 20%[IQR = 41] for LGA-SPM8, 14% [31] for LGA-SPM12 and 10% [27] with the LPA cross-sectional pipeline. Applying the LST longitudinal pipeline, the reproducibility errors were considerably reduced (LGA: 0%[IQR = 0], p < 0.001; LPA: 0% [3], p < 0.001) compared to those derived using the cross-sectional algorithms. The DC using the longitudinal pipeline was excellent (median = 1) for LGA [IQR = 0] and LPA [0.02]. LST algorithms showed moderate accuracy and good reproducibility. Therefore, it can be used as a reliable cross-sectional and longitudinal tool in multi-site studies.     

Cardiac magnetic resonance tissue tracking in right ventricle: Feasibility and normal values

PurposeTo investigate right ventricular (RV) strain in patients without identified cardiac pathology using cardiac magnetic resonance tissue tracking (CMR TT).MethodsA total of 50 consecutive patients with no identified cardiac pathology were analyzed. RV longitudinal and circumferential strain was assessed by CMR TT. The age range was 4–81 years with a median of 32 years (interquartile range, 15 to 56 years).ResultsAnalysis time per patient was < 5 min. The peak longitudinal strain (E_ll) was − 22.11 ± 3.51%. The peak circumferential strains (E_cc) for global, basal, mid-cavity and apical segments were as follows: − 11.69 ± 2.25%, − 11.00 ± 2.45%, − 11.17 ± 3.36%, − 12.90 ± 3.34%. There were significant gender differences in peak E_cc at the base (P = 0.04) and the mid-cavity (P = 0.03) with greater deformation in females than in males. On Bland-Altman analysis, peak E_ll (mean bias, 0.22 ± 1.67; 95% CI − 3.05 to 3.49) and mid-cavity E_cc (mean bias, 0.036 ± 1.75; 95% CI, − 3.39 to 3.47) had the best intra-observer agreement and inter-observer agreement, respectively.ConclusionsRV longitudinal and circumferential strains can be quickly assessed with good intra-observer and inter-observer variability using TT.     

Quantifying left atrial structure and function using single-plane tissue-tracking cardiac magnetic resonance

PurposeLeft atrial (LA) structure and function are important markers of adverse cardiovascular outcomes. Tissue-tracking cardiovascular magnetic resonance (CMR) accurately quantifies LA volume, strain, and strain rate based on biplane long-axis imaging. We aimed to assess the accuracy of the LA indices quantification from single-plane tissue-tracking CMR.MethodsWe included 388 subjects (mean age  57±13, male 70%) whose cine CMR images in sinus rhythm were available in both four-chamber and two-chamber views: 162 patients from the Prospective Observational Study of Implantable Cardioverter-Defibrillators (PROSE-ICD) Study, 208 patients from atrial fibrillation cohort, and 18 healthy volunteers. The group was divided into the training set (n = 291) and the test set (n = 97). In the training set, we compared the LA indices derived from biplane imaging and single-plane imaging (a four-chamber view), and developed regression equations. In the test set, we used the regression equations to estimate the LA indices from the single-plane imaging, and quantified the accuracy of the estimation against the LA indices from the biplane.ResultsIn the training set, all the LA indices from the single-plane imaging tended to be systematically underestimated compared with those from the biplane imaging, however, the correlation coefficient was high (r² = 0.73–0.90, p < 0.001). In the test set, LA volumetric indices showed excellent reproducibility (intra-class correlation coefficient (ICC): 0.91–0.92) with relatively low variability (16.3–22.3%); For LA strain and strain rate indices, reproducibility was excellent (ICC: 0.81–0.93), however, the variability was slightly higher than that of volumetric indices (21.7–25.4%).ConclusionsLA volumetric indices measured from single-plane tissue-tracking CMR are highly accurate and reproducible with reference to those derived from the standard biplane imaging. The reproducibility of LA strain and strain rate indices from single-plane tissue-tracking CMR is excellent but the variability is higher than that of the volumetric indices.     

Handcrafted MRI radiomics and machine learning: Classification of indeterminate solid adrenal lesions

PurposeTo assess a radiomic machine learning (ML) model in classifying solid adrenal lesions (ALs) without fat signal drop on chemical shift (CS) as benign or malignant.Method55 indeterminate ALs (21 lipid poor adenomas, 15 benign pheocromocytomas, 1 oncocytoma, 12 metastases, 6 primary tumors) showing no fat signal drop on CS were retrospectively included. Manual 3D segmentation on T2-weighted and CS images was performed for subsequent radiomic feature extraction. After feature stability testing and an 80–20% train-test split, the train set was balanced via oversampling. Following a multi-step feature selection, an Extra Trees model was tuned with 5-fold stratified cross-validation in the train set and then tested on the hold-out test set.ResultsA total of 3396 features were extracted from each AL, of which 133 resulted unstable while none had low variance (< 0.01). Highly correlated (r > 0.8) features were also excluded, leaving 440 parameters. Among these, Support Vector Machine 5-fold stratified cross-validated recursive feature elimination selected a subset of 6 features. ML obtained a cross-validation accuracy of 0.94 on the train and 0.91 on the test sets. Precision, recall and F1 score were respectively 0.92, 0.91 and 0.91.ConclusionsOur MRI handcrafted radiomics and ML pipeline proved useful to characterize benign and malignant solid indeterminate adrenal lesions.     

Cardiac MRI for the prognostication of heart failure with preserved ejection fraction: A systematic review and meta-analysis

BackgroundCardiac magnetic resonance imaging (MRI) is emerging as an important imaging tool in the assessment of heart failure with preserved ejection fraction (HFpEF). This systematic review and meta-analysis aim to synthesise and consolidate the current literature on cardiac MRI for prognostication of HFpEF.Methods designSystematic review and meta-analysis. Data sources: Scopus (PubMed and Embase) for studies published between 2008 and 2019. Eligibility criteria for study selection were studies that evaluated the prognostic role of cardiac MRI in HFpEF. Random effects meta-analyses of the reported hazard ratios (HR) for clinical outcomes was performed.ResultsInitial screening identified 97 studies. From these, only nine (9%) studies met all the criteria. The main cardiac MRI methods that demonstrated association to prognosis in HFpEF included late gadolinium enhancement (LGE) assessment of scar (n = 3), tissue characterisation with T1-mapping (n = 4), myocardial ischaemia (n = 1) and right ventricular dysfunction (RVSD) (n = 1). The pooled HR for all 9 studies was 1.52 (95% CI 1.05–1.99, P < 0.01). Sub-evaluation by cardiac MRI methods revealed varying HRs: LGE (net n = 402, HR = 1.6, 95% CI 0.42–2.78, P = 0.008); T1-mapping (n = 1623, HR = 1.25, 95% CI 0.891–1.60, P < 0.001); myocardial ischaemia or RVSD (n = 325, HR = 3.19, 95% CI 0.30–6.08, P = 0.03).ConclusionThis meta-analysis demonstrates that multiparametric cardiac MRI has value in prognostication of patients with HFpEF. HFpEF patients with a detectable scar on LGE, fibrosis on T1-mapping, myocardial ischaemia or RVSD appear to have a worse prognosis.PROSPERO registration numberCRD42020187228.     

Real-time cardiac magnetic resonance cine imaging with sparse sampling and iterative reconstruction for left-ventricular measures: Comparison with gold-standard segmented steady-state free precession

BackgroundSegmented cine imaging with a steady-state free-precession sequence (Cine-SSFP) is currently the gold standard technique for measuring ventricular volumes and mass, but due to multi breath-hold (BH) requirements, it is prone to misalignment of consecutive slices, time consuming and dependent on respiratory capacity. Real-time cine avoids those limitations, but poor spatial and temporal resolution of conventional sequences has prevented its routine application. We sought to examine the accuracy and feasibility of a newly developed real-time sequence with aggressive under-sampling of k-space using sparse sampling and iterative reconstruction (Cine-RT).MethodsStacks of short-axis cines were acquired covering both ventricles in a 1.5 T system using gold standard Cine-SSFP and Cine-RT. Acquisition parameters for Cine-SSFP were: acquisition matrix of 224 × 196, temporal resolution of 39 ms, retrospective gating, with an average of 8 heartbeats per slice and 1–2 slices/BH. For Cine-RT: acquisition matrix of 224 × 196, sparse sampling net acceleration factor of 11.3, temporal resolution of 41 ms, prospective gating, real-time acquisition of 1 heart-beat/slice and all slices in one BH. LV contours were drawn at end diastole and systole to derive LV volumes and mass.ResultsForty-one consecutive patients (15 male; 41 ± 17 years) in sinus rhythm were successfully included. All images from Cine-SSFP and Cine-RT were considered to have excellent quality. Cine-RT-derived LV volumes and mass were slightly underestimated but strongly correlated with gold standard Cine-SSFP. Inter- and intra-observer analysis presented similar results between both sequences.ConclusionsCine-RT featuring sparse sampling and iterative reconstruction can achieve spatial and temporal resolution equivalent to Cine-SSFP, providing excellent image quality, with similar precision measurements and highly correlated and only slightly underestimated volume and mass values.     

Quantitative assessment of regional systolic and diastolic functions and temporal heterogeneity of myocardial contraction in patients with myocardial infarction using cine magnetic resonance imaging and Fourier fitting

Purpose

The objective of this study is to determine regional left ventricle (LV) function and temporal heterogeneity of LV wall contraction by analyzing regional time–volume curve (TVC) after Fourier fitting and to assess altered systolic and diastolic functions and temporal indices of myocardial contraction in infarcted segments in comparison with noninfarcted myocardium in patients with myocardial infarction (MI).

Methods

Steady-state cine magnetic resonance (MR) and late gadolinium-enhanced (LGE) MR images were acquired using a 1.5-T MR system in 60 patients with MI. Regional LV function was determined by analyzing regional TVC in 16 segments. The fitted regional TVC was generated by Fourier curve fitting with five harmonics. Regional LV ejection fraction (EF), peak ejection rate (PER), peak filling rate (PFR), time to end-systole and time to peak filling (TPF) were determined from TVC and the first derivative curve.

Results

On LGE MR imaging (MRI), MI was observed in 307 of 960 segments (32.0%). Regional EF and PER averaged in LGE segments were 49.3±14.5% and 2.83±0.65 end-diastolic volume (EDV)/s, significantly lower than those in normal segments (66.7±11.9% and 3.63±0.60 EDV/s, P<.001 and P<.01, respectively). In addition, regional PFR, an index of diastolic function, was significantly reduced in LGE segments (1.94±0.54 vs. 2.86±0.68 EDV/s, P<.01). Time to end-systole and TPF were significantly greater in LGE segments (380.2±57.6 and 169.3±45.4 ms) than in normal segments (300.9±55.1 and 132.3±43.0 ms, P<.01 and P<.01, respectively).

Conclusions

Analysis of regional TVC on cine MRI after Fourier fitting allows quantitative assessment of regional systolic and diastolic LV functions and temporal heterogeneity of LV wall contraction in patients with MI.