CMR-based right ventricular strain analysis in cardiac amyloidosis and its potential as a supportive diagnostic feature.

Background: Right ventricular (RV) strain has provided valuable prognostic information for patients with cardiac amyloidosis (CA). However, the extent to which RV strain and strain rate can differentiate CA is not yet clinically established. CA underdiagnosis delays treatment strategies and exacerbates patient prognosis. Aims: Evaluation of cardiac magnetic resonance (CMR) quantified RV global and regional strain of CA and HCM patients along with CA subtypes. Methods: CMR feature tracking attained longitudinal, radial and circumferential global and regional strain in 47 control subjects (CTRL), 43 CA-, 20 hypertrophic cardiomyopathy- (HCM) patients. CA patients were subdivided in 21 transthyretin-related amyloidosis (ATTR) and 20 acquired immunoglobulin light chain (AL) patients. Strain data and baseline clinical parameters were statistically analysed with respect to diagnostic performance and discriminatory power between the different clinical entities. Results: Effective differentiation of CA from HCM patients was achieved utilizing global longitudinal (GLS: 16.5 ± 3.9% vs. -21.3 ± 6.7%, p = 0.032), radial (GRS: 11.7 ± 5.3% vs. 16.5 ± 7.1%, p < 0.001) and circumferential (GCS: -7.6 ± 4.0% vs. -9.4 ± 4.4%, p = 0.015) right ventricular strain. Highest strain-based hypertrophic phenotype differentiation was attained using GRS (AUC = 0.86). Binomial regression found right ventricular ejection fraction (RV-EF) (p = 0.017) to be a significant predictor of CA-HCM differentiation. CA subtypes had comparable cardiac strains. Conclusion: CMR-derived RV global strains and various regional longitudinal strains provide discriminative radiological features for CA-HCM differentiation. However, in terms of feasibility, cine-derived RV-EF quantification may suffice for efficient differential diagnostic support.

A Machine Learning Challenge: Detection of Cardiac Amyloidosis Based on Bi-Atrial and Right Ventricular Strain and Cardiac Function.

BACKGROUND: This study challenges state-of-the-art cardiac amyloidosis (CA) diagnostics by feeding multi-chamber strain and cardiac function into supervised machine (SVM) learning algorithms. METHODS: Forty-three CA (32 males; 79 years (IQR 71; 85)), 20 patients with hypertrophic cardiomyopathy (HCM, 10 males; 63.9 years (±7.4)) and 44 healthy controls (CTRL, 23 males; 56.3 years (IQR 52.5; 62.9)) received cardiovascular magnetic resonance imaging. Left atrial, right atrial and right ventricular strain parameters and cardiac function generated a 41-feature matrix for decision tree (DT), k-nearest neighbor (KNN), SVM linear and SVM radial basis function (RBF) kernel algorithm processing. A 10-feature principal component analysis (PCA) was conducted using SVM linear and RBF. RESULTS: Forty-one features resulted in diagnostic accuracies of 87.9% (AUC = 0.960) for SVM linear, 90.9% (0.996; Precision = 94%; Sensitivity = 100%; F1-Score = 97%) using RBF kernel, 84.9% (0.970) for KNN, and 78.8% (0.787) for DT. The 10-feature PCA achieved 78.9% (0.962) via linear SVM and 81.8% (0.996) via RBF SVM. Explained variance presented bi-atrial longitudinal strain and left and right atrial ejection fraction as valuable CA predictors. CONCLUSION: SVM RBF kernel achieved competitive diagnostic accuracies under supervised conditions. Machine learning of multi-chamber cardiac strain and function may offer novel perspectives for non-contrast clinical decision-support systems in CA diagnostics.

Left-Ventricular Reference Myocardial Strain Assessed by Cardiovascular Magnetic Resonance Feature Tracking and fSENC-Impact of Temporal Resolution and Cardiac Muscle Mass.

Aims: Cardiac strain parameters are increasingly measured to overcome shortcomings of ejection fraction. For broad clinical use, this study provides reference values for the two strain assessment methods feature tracking (FT) and fast strain-encoded (fSENC) cardiovascular magnetic resonance (CMR) imaging, including the child/adolescent group and systematically evaluates the influence of temporal resolution and muscle mass on strain. Methods and Results: Global longitudinal (GLS), circumferential (GCS), and radial (GRS) strain values in 181 participants (54% women, 11-70 years) without cardiac illness were assessed with FT (CVI42® software). GLS and GCS were also analyzed using fSENC (MyoStrain® software) in a subgroup of 84 participants (60% women). Fourteen patients suffering hypertrophic cardiomyopathy (HCM) were examined with both techniques. CMR examinations were done on a 3.0T MR-system. FT-GLS, FT-GCS, and FT-GRS were -16.9 ± 1.8%, -19.2 ± 2.1% and 34.2 ± 6.1%. fSENC-GLS was higher at -20.3 ± 1.8% (p < 0.001). fSENC-GCS was comparable at-19.7 ± 1.8% (p = 0.06). All values were lower in men (p < 0.001). Cardiac muscle mass correlated (p < 0.001) with FT-GLS (r = 0.433), FT-GCS (r = 0.483) as well as FT-GRS (r = -0.464) and acts as partial mediator for sex differences. FT-GCS, FT-GRS and fSENC-GLS correlated weakly with age. FT strain values were significantly lower at lower cine temporal resolutions, represented by heart rates (r = -0.301, -0.379, 0.385) and 28 or 45 cardiac phases per cardiac cycle (0.3-1.9% differences). All values were lower in HCM patients than in matched controls (p < 0.01). Cut-off values were -15.0% (FT-GLS), -19.3% (FT-GCS), 32.7% (FT-GRS), -17.2% (fSENC-GLS), and -17.7% (fSENC-GCS). Conclusion: The analysis of reference values highlights the influence of gender, temporal resolution, cardiac muscle mass and age on myocardial strain values.