Identifying Unique Patterns of Myocardial Deformation through Segmental Speckle Tracking Stress Strain Following High‐Fat Diet

Abstract

Early detection of functional and structural changes to the heart are pivotal in treating cardiovascular disease and mitigating morbidity and mortality. This study utilized segmental stress strain analysis to evaluate whether different insults (diet induced obesity (DIO) and chronic stress) cause preferential dysfunction to specific regions of the heart. Male C57Bl/6 mice were randomized into lean or obese groups (N = 24/group) at 6 weeks of age. Mice in the obese group underwent 12 weeks of DIO with a high-fat diet (HFD). At 18 weeks, mice in lean and obese groups were further randomized into unpredictable chronic mild stress (UCMS) and non-UCMS groups to elicit a chronically stressed phenotype (UCMS; 7 hrs/day, 5 days/week, for 8 weeks). M-Mode, Pulse-Wave Doppler, and speckle tracking stress-strain echocardiographic parameters were measured at baseline (6 weeks), post-HFD (17 weeks), and post-UCMS (26 weeks). Supervised machine learning identified the predictive capacity of each diagnostic echocardiographic technique in DIO and chronic stress using 10-fold cross validation. Comparing the DIO group to control mice, the machine learning model provided robust prediction (AUC: 0.921), with the two most important features being radial strain of the lateral wall and anterior free wall. Changes in radial strain of the lateral wall (-64%, P≤0.001) and anterior free wall (-53%, P<0.001) were more pronounced compared to ejection fraction (-9%, P<0.001) and global longitudinal strain (-16%, P=0.08). The ability to predict mice that underwent chronic stress, irrespective of diet, was assessed (AUC: 0.714) revealing longitudinal strain rate of the anterior mid wall and radial strain of the posterior septal wall as the top two features. The wall segments indicate a predilection for changes in deformation patterns to the free wall (DIO) and septal wall (chronic stress). This study suggests that segmental analysis can provide early indications of disease-specific changes to the myocardium.