Discrimination between the presence and absence of spontaneous circulation using smartphone seismocardiography: A preliminary investigation

Background: Seismocardiography measures the vibrations produced by the beating heart using an accelerometer sensor placed on the chest. We evaluated the ability of smartphone seismocardiography to distinguish between the presence and absence of spontaneous circulation.

Methods: Seismocardiography signals were obtained using a smartphone placed on the sternum in a convenience sample of 60 adult patients (30 comatose patients with spontaneous circulation and 30 deceased patients). The maximum, minimum, and standard deviation (SD) of acceleration values for head-to-foot, right-to-left, and dorsoventral axes and the three axis-root mean square (RMS) of the acceleration signals were calculated. Blinded observers (n = 156) were each asked to determine the presence or absence of spontaneous circulation based on seismocardiography video clips for each of the 60 patients.

Results: The seismocardiography revealed periodic large positive peaks in the patients with spontaneous circulation, which were absent in the patients without spontaneous circulation. For each of the four output measurements (three independent axes plus the three-axis RMS), the acceleration maxima and SD were significantly higher and the minima significantly lower in the patients with spontaneous circulation than in those without spontaneous circulation (all P < 0.001 except the minimum of three axis-RMS results [P = 0.009]). The observers accurately identified the seismocardiography signals from patients without spontaneous circulation, with a sensitivity of 97.6% (95% confidence interval, 97.0%-98.2%) and a specificity of 98.4% (95% confidence interval, 97.8%-99.0%).

Conclusions: In conclusion, blinded observers accurately distinguished between seismocardiography signals from patients with and without spontaneous circulation.