Myocardial Function Prediction After Coronary Artery Bypass Grafting Using MRI Radiomic Features and Machine Learning Algorithms.

The main aim of the present study was to predict myocardial function improvement in cardiac MR (LGE-CMR) images in patients after coronary artery bypass grafting (CABG) using radiomics and machine learning algorithms. Altogether, 43 patients who had visible scars on short-axis LGE-CMR images and were candidates for CABG surgery were selected and enrolled in this study. MR imaging was performed preoperatively using a 1.5-T MRI scanner. All images were segmented by two expert radiologists (in consensus). Prior to extraction of radiomics features, all MR images were resampled to an isotropic voxel size of 1.8 × 1.8 × 1.8 mm3. Subsequently, intensities were quantized to 64 discretized gray levels and a total of 93 features were extracted. The applied algorithms included a smoothly clipped absolute deviation (SCAD)-penalized support vector machine (SVM) and the recursive partitioning (RP) algorithm as a robust classifier for binary classification in this high-dimensional and non-sparse data. All models were validated with repeated fivefold cross-validation and 10,000 bootstrapping resamples. Ten and seven features were selected with SCAD-penalized SVM and RP algorithm, respectively, for CABG responder/non-responder classification. Considering univariate analysis, the GLSZM gray-level non-uniformity-normalized feature achieved the best performance (AUC: 0.62, 95% CI: 0.53-0.76) with SCAD-penalized SVM. Regarding multivariable modeling, SCAD-penalized SVM obtained an AUC of 0.784 (95% CI: 0.64-0.92), whereas the RP algorithm achieved an AUC of 0.654 (95% CI: 0.50-0.82). In conclusion, different radiomics texture features alone or combined in multivariate analysis using machine learning algorithms provide prognostic information regarding myocardial function in patients after CABG.

Correlation of semi-quantitative findings of endolymphatic hydrops in MRI with the audiometric findings in patients with Meniere's disease.

Purpose: To investigate the correlation between vestibular hydrops (VH), cochlearhydrops (CH), vestibular aqueduct non-visibility (VANV), and visually increased perilymphatic enhancement (VIPE) with the findings of pure-tone audiometry (PTA) in Meniere's disease (MD) patients. Methods: In this cross-sectional study, 53 ears belonging to 48 patients were divided into two groups and evaluated. In group "MD patients," there were 24 ears of 19 patients diagnosed with the definite MD (14 patients with unilateral and 5 patients withbilateral involvements). The "control group" consisted of 29 non-symptomatic ears belonging to patients diagnosed with unilateral sudden sensory-neural hearing loss or unilateral schwannoma. All the patients underwent 2 sessions of temporal bone MRI using the same 3T system: an unenhanced axial T1, T2, and 3D-FLAIR MRI, an intravenous gadolinium-enhanced axial T1 fat-sat, and 4 h after the injection, an axial 3D-T2 cube and 3D-FLAIR session. VH, CH, VANV, and VIPE were assessed. Subsequently, the correlation between EH indices and PTA findings (in three frequency domains of low, middle, and high) were evaluated, and the predictive value of MRI was calculated. Results: VH was significantly correlated with the hearing threshold in the low, middle, and high-frequency domains. CH was also correlated with the hearing threshold in the low and middle domains. Contrarily, VIPE was not associated with hearing thresholds, and VANV was only correlated with the hearing threshold in low frequencies. Conclusion: The grade of VH, CH, and VANV were significantly correlated with the hearing thresholds in PTA.

Non-contrast Cine Cardiac Magnetic Resonance image radiomics features and machine learning algorithms for myocardial infarction detection.

OBJECTIVE: Robust differentiation between infarcted and normal tissue is important for clinical diagnosis and precision medicine. The aim of this work is to investigate the radiomic features and to develop a machine learning algorithm for the differentiation of myocardial infarction (MI) and viable tissues/normal cases in the left ventricular myocardium on non-contrast Cine Cardiac Magnetic Resonance (Cine-CMR) images. METHODS: Seventy-two patients (52 with MI and 20 healthy control patients) were enrolled in this study. MR imaging was performed on a 1.5 T MRI using the following parameters: TR = 43.35 ms, TE = 1.22 ms, flip angle = 65°, temporal resolution of 30-40 ms. N4 bias field correction algorithm was applied to correct the inhomogeneity of images. All images were segmented and verified simultaneously by two cardiac imaging experts in consensus. Subsequently, features extraction was performed within the whole left ventricular myocardium (3D volume) in end-diastolic volume phase. Re-sampling to 1 × 1 × 1 mm3 voxels was performed for MR images. All intensities within the VOI of MR images were discretized to 64 bins. Radiomic features were normalized to obtain Z-scores, followed by Student's t-test statistical analysis for comparison. A p-value < 0.05 was used as a threshold for statistically significant differences and false discovery rate (FDR) correction performed to report q-value (FDR adjusted p-value). The extracted features were ranked using the MSVM-RFE algorithm, then Spearman correlation between features was performed to eliminate highly correlated features (R2 > 0.80). Ten different machine learning algorithms were used for classification and different metrics used for evaluation and various parameters used for models' evaluation. RESULTS: In univariate analysis, the highest area under the curve (AUC) of receiver operating characteristic (ROC) value was achieved for the Maximum 2D diameter slice (M2DS) shape feature (AUC = 0.88, q-value = 1.02E-7), while the average of univariate AUCs was 0.62 ± 0.08. In multivariate analysis, Logistic Regression (AUC = 0.93 ± 0.03, Accuracy = 0.86 ± 0.05, Recall = 0.87 ± 0.1, Precision = 0.93 ± 0.03 and F1 Score = 0.90 ± 0.04) and SVM (AUC = 0.92 ± 0.05, Accuracy = 0.85 ± 0.04, Recall = 0.92 ± 0.01, Precision = 0.88 ± 0.04 and F1 Score = 0.90 ± 0.02) yielded optimal performance as the best machine learning algorithm for this radiomics analysis. CONCLUSION: This study demonstrated that using radiomics analysis on non-contrast Cine-CMR images enables to accurately detect MI, which could potentially be used as an alternative diagnostic method for Late Gadolinium Enhancement Cardiac Magnetic Resonance (LGE-CMR).

Myocardial strain by cardiac magnetic resonance: A valuable predictor of outcome after infarct revascularization.

PURPOSE: To evaluate the prognostic value of left ventricular strains by cardiac magnetic resonance feature tracking (CMR-FT) in patients with re-perfused myocardial infarction (MI). METHODS: The study enrolled 58 patients with re-vascularized MI who underwent CMR within a week from acute MI. An 18-month follow-up was carried out for the composite endpoint of major adverse cardiovascular events (MACE). A 3 to 6-month post-MI ejection fraction (EF) was also measured. The predictive value of global longitudinal, circumferential, and radial strains (GLS, GCS, and GRS, respectively) for MACE and the follow-up EF was evaluated. RESULTS: All the global strains showed significant impairment in MACE positive cases (P < 0.05 for all). On univariate regression, MACE was reversely associated with early post-MI EF (OR: 0.90, 95% CI: 0.83-0.98, P: 0.01), and directly associated with GLS (OR: 1.32, 95% CI: 1.03-1.69, P: 0.02), GCS (OR: 1.23, 95% CI: 1.00-1.50, P: 0.04) and EDVI (OR:1.02, 95 %CI: 1.00-1.04, P: 0.01). On multivariate regression model, only the interaction between EF and GLS showed a significant association with MACE (OR[CI95%]: 1.1 [1.06-1.21]). EF < 30% and GLS > -8.9% had the highest sensitivity (78.9% and 89.5%, respectively) and specificity (45.2% and 54.8%, respectively) to predict MACE. The combination of EF < 30% and GLS > -8.9% increased the sensitivity to 94.7%. In addition, the cutoff values of 35.1% for early post-MI EF and -10% for GLS could identify patients with impaired follow-up EF with more than 80% sensitivity and specificity [AUC (CI95%): 0.893(0.76-1.00) for EF and AUC (CI95%):0.836(0.67-1,00) for GLS, P < 0.05 for both)]. CONCLUSIONS: GLS by CMR-FT is a powerful prognosticator of MACE and functional recovery in MI survivors, with incremental value added to early post-MI EF alone.

Efficacy of Novel Noncontrast Cardiac Magnetic Resonance Methods in Indicating Fibrosis in Hypertrophic Cardiomyopathy.

OBJECTIVE: In hypertrophic cardiomyopathy (HCM), myocardial fibrosis is routinely shown by late gadolinium enhancement (LGE) in cardiac magnetic resonance (CMR) imaging. We evaluated the efficacy of 2 novel contrast-free CMR methods, namely, diffusion-weighted imaging (DWI) and feature-tracking (FT) method, in detecting myocardial fibrosis. METHODS: This cross-sectional study was conducted on 26 patients with HCM. Visual and quantitative comparisons were made between DWI and LGE images. Regional longitudinal, circumferential, and radial strains were compared between LGE-positive and LGE-negative segments. Moreover, global strains were compared between LGE-positive and LGE-negative patients as well as between patients with mild and marked LGE. RESULTS: All 3 strains showed significant differences between LGE-positive and LGE-negative segments (P < 0.001). The regional longitudinal and circumferential strain parameters showed significant associations with LGE (P < 0.001), while regional circumferential strain was the only independent predictor of LGE in logistic regression models (OR: 1.140, 95% CI: 1.073 to 1.207, P < 0.001). A comparison of global strains between patients with LGE percentages of below 15% and above 15% demonstrated that global circumferential strain was the only parameter to show impairment in the group with marked myocardial fibrosis, with borderline significance (P=0.09). A review of 212 segments demonstrated a qualitative visual agreement between DWI and LGE in 193 segments (91%). The mean apparent diffusion coefficient was comparable between LGE-positive and LGE-negative segments (P=0.51). CONCLUSIONS: FT-CMR, especially regional circumferential strain, can reliably show fibrosis-containing segments in HCM. Further, DWI can function as an efficient qualitative method for the estimation of the fibrosis extent in HCM.