Fractal analysis of left ventricular trabeculae in hypertensive patients with heart failure: a 3.0 T cardiac magnetic resonance study.

Background: Endocardial trabecular hyperplasia due to hemodynamic stress reflects phenotypic variability in disease progression. Employing fractal analysis, this study quantified left ventricular (LV) myocardial trabecular complexity in hypertensive patients with and without heart failure (HF) to evaluate its diagnostic utility for HF. Methods: This study retrospectively enrolled 146 hypertensive patients (77 with HF, 69 without), grouped into HTN-HF (n = 77) and HTN non-HF (n = 69); additionally, 34 healthy controls were recruited. Clinical data and cardiac MRI parameters were compared. Fractal dimension (FD) values were calculated on the LV short-axis cine images using fractal analysis. Logistic regression analysis was performed to determine predictors. Results: Five fractal dimensions were derived: global FD, along with mean/maximal apical FD and mean/maximal basal FD. Compared with healthy controls, HF patients showed significantly elevated left ventricular fractal dimensions (all P < 0.001). Moreover, these fractal dimensions exhibited significant differences between the HTN-HF patients and HTN non-HF patients, except for maximal basal FD. The univariate logistic regression revealed that global FD, mean/maximal apical FD and mean basal FD emerged as significant independent predictors (OR: 1.170,1.121,1.070, and 1.088, P < 0.05). Furthermore, integration of fractal dimensions enhanced calibration and diagnostic accuracy of the model. (AUC: 0.877). Conclusions: CMR fractal analysis provides a feasible technique for quantifying LV myocardial trabecular complexity in hypertensive heart failure patients. In conclusion, our study demonstrates the potential of fractal analysis to provide incremental diagnostic value for heart failure within the hypertensive population. Integration of FD into clinical diagnostic models may enhance diagnostic performance.