Clinical Readiness of Hemodynamic Estimation from Cardiac Vibration Signals: An Endpoint-Specific Critical Review.

purposeCardiac vibration signals have been investigated as unobtrusive surrogates of central hemodynamics, but the literature frequently mixes physiological plausibility, association, and clinically meaningful validation. This review evaluates the validation readiness of vibration-based estimation of central hemodynamic parameters, specifically stroke volume (SV), cardiac output (CO), and arterial blood pressure (BP), including cuffless or central BP claims when explicitly reported. Ejection fraction (EF) and pulmonary capillary wedge pressure (PCWP) are also considered as related endpoints for cardiac function and filling pressure.

methodsA structured narrative review was conducted across studies using seismocardiography, ballistocardiography, gyrocardiography, and kinocardiography. Evidence was appraised by endpoint using reference standard tier, agreement analysis, calibration burden, and population realism, with emphasis on translational interpretability rather than correlation alone.

resultsEvidence maturity differed substantially by endpoint. The strongest support was found for stroke volume and cardiac output. For these endpoints, weighing-scale ballistocardiography and wearable seismocardiography systems validated against catheterization-based Direct Fick and thermodilution measurements provided the clearest agreement-based evidence within the studies reviewed. Clinically meaningful accuracy remains use-case dependent. Ejection fraction studies were dominated by classification and biomarker-style analyses rather than continuous estimation with agreement reporting. Exploratory early diastolic SCG studies remain limited to associations with echocardiographic relaxation indices, such as tissue Doppler e', and preload-perturbation responses, rather than direct preload measurement or validated diastolic-function estimation. Blood pressure studies reported promising calibrated error estimates in small healthy cohorts, but none met the International Organization for Standardization (ISO) 81060-2 requirements.

conclusionCardiac vibration sensing is not uniformly ready for clinical translation. Stroke volume and cardiac output estimation currently have a relatively more developed validation foundation, whereas ejection fraction and cuffless blood pressure remain pre-validation endpoints that require larger, more clinically realistic studies and stronger agreement-based evaluation.