Hypertrophic cardiomyopathy (HCM) is a relatively common inherited cardiac disease that results in left ventricular hypertrophy. Machine learning uses algorithms to study patterns in data and develop models able to make predictions. The aim of this study is to identify HCM subtypes and examine the mechanisms of HCM using machine learning algorithms. Clinical and laboratory findings of 143 adult patients with a confirmed diagnosis of nonobstructive HCM are analyzed; HCM subtypes are determined by clustering, while the presence of different HCM features is predicted in classification machine learning tasks. Four clusters are determined as the optimal number of clusters for this dataset. Models that can predict the presence of particular HCM features from other genotypic and phenotypic information are generated, and subsets of features sufficient to predict the presence of other features of HCM are determined. This research proposes four subtypes of HCM assessed by machine learning algorithms and based on the overall phenotypic expression of the participants of the study. The identified subsets of features sufficient to determine the presence of particular HCM aspects could provide deeper insights into the mechanisms of HCM.

Purpose The present study i) determined left atrial (LA) and left ventricular (LV) strains at rest and in response to exercise in patients with heart failure with preserved ejection fraction (HFpEF) and heart failure with reduced ejection fraction (HFrEF), and ii) assessed the relationship between LA and LV strains and exercise tolerance. Methods Forty HFpEF patients (age 59±7 yrs; 25 females), 40 stable HFrEF patients (age 57+6 yrs, 15 females) and 20 controls (age 56+6 yrs, 13 females) underwent baseline clinical and biochemical assessment, resting and exercise stress transthoracic echocardiography using modified Bruce protocol. Speckle-tracking echocardiography was performed to define peak atrial longitudinal strain (PALS) and left ventricular global longitudinal strain (LVGLS). LA stiffness index and LV stiffness index were also derived. Results Compared to healthy controls, HFpEF and HFrEF showed significantly lower PALS at rest (34.03±1.85 vs. 23.06±4.69 vs. 11.51±1.44%, p<0.01) and after exercise (34.41±1.24 vs. 18.48±3.51 vs 10.47±1.49, p<0.01 for both). In response to exercise, the PALS significantly reduced in HFpEF by 26%, but only 8% in HFrEF and remained unchanged in controls. LA stiffness index was higher in HFpEF and HFrEF compared to healthy controls at rest (0.57±0.22 vs. 1.19±0.63 vs. 0.27±0.06, p<0.01) and in response to exercise (0.83±0.46 vs. 1.37±0.63 vs. 0.33±0.04, p<0.01). Compared to healthy controls, HFpEF and HFrEF demonstrated significantly lower LVGLS at rest (−20.27±0.98 vs. −15.89±2.72 vs.-11.14±3.40%, p<0.01) and after exercise (−19.9±0.8 vs.-15.5±3.18 vs.-11.01±2.6%, p<0.01). LV stiffness index was significantly higher in HFpEF and HFrEF compared to healthy controls at rest (0.16±0.05 vs. 0.14±0.07 vs. 0.11±0.02, p<0.01) and in response to exercise (0.18±0.07 vs. 0.15±0.06 vs. 0.13±0.02, p<0.01). Exercise tolerance i.e. exercise duration was significantly lower by 28% and 30% in HFpEF and HFrEF compared with controls (363±152 vs. 352±91 vs. 505±42, p<0.01). There was a significant relationship between peak atrial longitudinal strain and exercise tolerance in HFpEF (r=0.32, p=0.04).There was no significant relationship between exercise tolerance and LVGLS (r=0.058, p=0.72), LA stiffness (r=−0.17, p=0.3), LV stiffness (r=0.88, p=.59). There was no significant relationship between exercise tolerance and PALS (r=0.021, p=0.89) or LVGLS (r=0.12, p=0.48) in HFrEF. Conclusion HFpEF and HFrEF are associated with reduced left atrial and left ventricular strains and increased arial and ventricular stiffness.Peak atrial longitudinal strain is a significant determinant of exercise tolerance in HFpEF but not in HFrEF. Funding Acknowledgement Type of funding sources: None.