Heart Rate Monitoring During Different Lung Volume Phases Using Seismocardiography

Seismocardiography (SCG) is a non-invasive method that can be used for cardiac activity monitoring. This paper presents a new electrocardiogram (ECG) independent approach for estimating heart rate (HR) during low and high lung volume (LLV and HLV, respectively) phases using SCG signals. In this study, SCG, ECG, and respiratory flow rate (RFR) signals were measured simultaneously in 7 healthy subjects. The lung volume information was calculated from the RFR and was used to group the SCG events into low and high lung-volume groups. LLV and HLV SCG events were then used to estimate the subjects HR as well as the HR during LLV and HLV in 3 different postural positions, namely supine, 45 degree heads-up, and sitting. The performance of the proposed algorithm was tested against the standard ECG measurements. Results showed that the HR estimations from the SCG and ECG signals were in a good agreement (bias of 0.08 bpm). All subjects were found to have a higher HR during HLV (HR$\text{HLV}$) compared to LLV (HR$\text{LLV}$) at all postural positions. The HR$\text{HLV}$/HR$\text{LLV}$ ratio was 1.11$\pm$0.07, 1.08$\pm$0.05, 1.09$\pm$0.04, and 1.09$\pm$0.04 (mean$\pm$SD) for supine, 45 degree-first trial, 45 degree-second trial, and sitting positions, respectively. This heart rate variability may be due, at least in part, to the well-known respiratory sinus arrhythmia. HR monitoring from SCG signals might be used in different clinical applications including wearable cardiac monitoring systems.

💡 Research Summary

The paper introduces an ECG‑independent method for estimating heart rate (HR) during distinct phases of the respiratory cycle—low lung volume (LLV) and high lung volume (HLV)—by exploiting seismocardiography (SCG) signals. Seven healthy young male volunteers were studied while simultaneously recording SCG, electrocardiogram (ECG), and respiratory flow rate (RFR) at a high sampling rate (10 kHz, down‑sampled to 320 Hz). The RFR signal was integrated to obtain a lung‑volume waveform, which was then used to label each SCG cardiac event as belonging to either the LLV or HLV phase.

The signal‑processing pipeline began with low‑pass filtering of SCG (cut‑off 100 Hz) and peak detection to isolate individual cardiac events. A simple pseudo‑code algorithm grouped events based on the sign of the lung‑volume signal: positive values indicated HLV, negative values indicated LLV. For each group, the interval between successive events was converted to an instantaneous HR (HR = 60 / Δt). To avoid spurious low‑rate estimates caused by events that straddle two respiratory cycles, any HR below 50 bpm was discarded. The final HR for a recording was computed as a weighted average of the LLV‑derived and HLV‑derived HR values, using the number of events in each group as weights (Equation 2).

Performance was benchmarked against the ECG‑derived HR, which served as the gold standard. Bland‑Altman analysis revealed a mean bias of only 0.08 bpm, with 95 % limits of agreement tightly clustered around zero, indicating excellent concordance between the SCG‑based and ECG‑based measurements. Across three postural conditions—supine, 45° head‑up tilt (two trials), and sitting—all subjects exhibited higher HR during HLV than during LLV. The ratio HR_HLV / HR_LLV was 1.11 ± 0.07 (supine), 1.08 ± 0.05 (45° trial 1), 1.09 ± 0.04 (45° trial 2), and 1.09 ± 0.04 (sitting). These findings are consistent with respiratory sinus arrhythmia (RSA), a well‑known physiological phenomenon whereby heart rate accelerates during inspiration (high lung volume) and decelerates during expiration (low lung volume) to improve ventilation‑perfusion matching.

The authors discuss the clinical relevance of being able to detect respiration‑linked HR changes without ECG. Since SCG can be captured with a simple accelerometer, the technique is well suited for wearable devices that aim to provide continuous cardiac monitoring in everyday life. Potential applications include early detection of cardiovascular dysfunction, remote patient monitoring, and real‑time HR feedback during exercise.

Limitations of the study are acknowledged: the sample size is small, all participants are male, and the breathing pattern was controlled using a volume‑controlled ventilator, which may not fully represent spontaneous breathing in real‑world settings. Future work should involve larger, more diverse cohorts, free‑breathing protocols, and validation of the algorithm in ambulatory environments.

In conclusion, the paper demonstrates that lung‑volume‑segmented SCG events can yield accurate, ECG‑independent heart‑rate estimates and that the method reliably captures the physiological HR modulation associated with respiration. This advances the feasibility of SCG‑based wearable cardiac monitoring systems and opens avenues for integrating respiratory phase information into cardiovascular diagnostics.