Forcecardiography as a non-invasive predictor of echocardiographic parameters in patients undergoing cardiac resynchronization therapy: a preliminary study
A Salucci, A Simonetti, S Piccirillo, J Centracchio, E Andreozzi, S Parlato, V Pergola, G Ammirati, L Cocchiara, P Visconti, A Volpe, B Brescia, P Vergara, A Rapacciuolo, T StrisciuglioAbstract
Introduction
Forcecardiography (FCG) is a non-invasive technique that measures chest wall vibrations generated by cardiac mechanical activity, providing valuable insights into cardiac function. By detecting mechanical cardiac events such as valve opening and closure, FCG can complement ECG data to estimate key hemodynamic parameters. Identifying correlations between FCG-derived features and echocardiographic or electrocardiographic parameters may enable non-invasive assessment of cardiac performance in patients with heart failure.
Purpose
This study aimed to evaluate the relationship between FCG-derived features and ECG/ECHO parameters in patients with heart failure and CRT-D devices, under different pacing modalities.
Methods
FCG and ECG signals were recorded in 11 HFrEF patients (ischemic and non-ischemic etiologies) with CRT-D devices during intrinsic rhythm, biventricular (BiV), and right ventricular (RV) pacing. FCG features (energy, skewness, kurtosis) were compared with ECG parameters (e.g., QRS duration) and echocardiographic indices including stroke volume (SV), global longitudinal strain (GLS), LVOT velocity-time integral (VTI), and global work index (GWI). Regression analyses were used to assess predictive correlations, evaluated with R² and RMSE (Root Mean Squared Error).
Results
Significant correlations were found between FCG-derived features and ECG/ECHO parameters (coefficients: -0.53 to 0.85). Regression models demonstrated strong predictive performance for GWI (R² = 0.93, RMSE = 174.44 mmHg%), LVOT VTI (R² = 0.90, RMSE = 3.51 cm), and SV (R² = 0.91, RMSE = 12.09 mL).
Conclusions
FCG shows promise as a non-invasive tool for estimating echocardiographic metrics such as GWI, LVOT VTI, and SV. Among extracted features, kurtosis exhibited strong predictive value across both intrinsic rhythm and pacing conditions. Larger studies are needed to confirm these preliminary findings and support broader clinical application.