Past, Present, and Future of Evaluating Diastolic Function

The assessment of left ventricular diastolic function started with invasive pressure measurements and is currently primarily based on echocardiographic imaging. The current approach to the diagnosis of diastolic function relies on mitral inflow velocities, tissue Doppler early diastolic velocity of the mitral annulus, peak velocity of tricuspid regurgitation, pulmonary vein flow, left atrial size and function, and the noninvasive estimation of right atrial pressure. The recommended algorithm in the 2025 American Society of Echocardiography guidelines for estimation of mean left atrial pressure has been validated in a large multicenter study that included 951 patients. Further, that algorithm has shown incremental value to triaging clinical scores in achieving accurate heart failure with preserved ejection fraction diagnosis, against the invasive gold standard. Newer promising approaches have been developed over the past few years to study left ventricular diastolic function, including shear wave propagation velocity, left atrial conduit volume, and timing of mitral and tricuspid valve opening in the apical 4-chamber view. Artificial intelligence has been applied to this field as a rule-based decision tree algorithm, and as deep-learning neural networks to train models for diagnosing and grading diastolic dysfunction, and for diagnosing heart failure with preserved ejection fraction.