Assessment of Left Ventricular Filling Pressure by Applying Artificial Intelligence Algorithms to Left Atrial Speckle-tracking Echocardiography

Valutazione Delle Pressioni di Riempimento Del Ventricolo Sinistro Attraverso l'Impiego di un Software di Intelligenza Artificiale Applicato Alla Stima Dello Strain Atriale Sinistro Con Ecocardiografia Speckletracking: Studio di Validazione Mediante Cateterismo Cardiaco

The goal of this single-center, prospective, observational study is to validate a software package based on an artificial intelligence algorithm for automated non-invasive estimation of LVFP against invasively measured ones by left and right heart catheterization; In addition, the added value of this new automated software to detect increased LVEDP > 15 mm Hg compared to the current ASE/EACVI algorithm for assessing diastolic dysfunction and longitudinal left atrial strain during the reservoir phase will be evaluated.

Inclusion Criteria:

• ≥ 18 yo,
• Signed informed consent to be part of this study,
• No pregnancy,
• Good acoustic window and patient cooperation to obtain good 2- and 3-dimensional datasets focused on left atrium (LA),

Exclusion Criteria:

• Unwillingness to be part of the study,
• Patients with poor 2-dimensional imaging quality in ≥ 1 LA segment or lacking of a discernible R-R interval, which might preclude accurate strain measurements,
• Patients with more than mild mitral valve heart disease, aortic regurgitation, left side valvular prosthesis or mitral valve repair, mitral annulus calcification (≥ 5 mm),
• Patients with history or presence of atrial fibrillation (AF), supraventricular arrhythmias, ventricular paced rhythm, left bundle branch block (LBBB),
• Patients taking diuretics or vasodilators (nitrates) on the day of the examination, which could alter loading conditions.