Estimation of Sleep Onset Latency Based on the Blood Pressure Regulatory Reflex Mechanism

Abstract

Sleep onset latency (SOL) is an objective indicator of sleepiness and is being used clinically as a diagnostic tool for sleep deprivation. This study proposes a new and less intrusive approach to estimate SOL based on the blood pressure (BP) regulatory reflex mechanism.We hypothesized that the arterial baroreflex, one kind of reflex mechanism for BP regulation, maintains the toning-down effect sleep has on BP. The arterial baroreflex is strongly activated after the time of sleep onset in order to maintain the lowered BP by leading to an increase in heart beat interval (HBI). This observation suggests that the arterial baroreflex has a marked influence on the HBI control with the onset of sleep. As a result, a positive correlation is expected between fluctuations of BP and those of subsequent HBI after sleep onset. To investigate our hypothesis, we determined the relationship between BP and HBI using the R-J and R-R intervals measured from an electrocardiogram and a ballistocardiogram.We estimated SOL using the correlation coefficients corresponding to the relationship between fluctuations of the R-J interval and those of the subsequent R-R interval. The SOLs of ten healthy subjects [age (mean ± SD): 28.72 ± 3.21 years] were estimated using our proposed method and compared with the polysomnography data. The mean absolute error was 0.25 ± 0.35 min, corresponding to a half epoch (15 s) on average. We expect our method will be applicable as a nonintrusive and automatic SOL estimation system that does not require the use of electroencephalogram sensors.

Oxygen Desaturation Index Estimation through Unconstrained Cardiac Sympathetic Activity Assessment Using Three Ballistocardiographic Systems

“…Piezoelectric sensor-based BCG monitoring systems were used for the unconstrained acquisition of N-N intervals. With a growing need for long-term monitoring techniques in sleep studies, ballistocardiography has been suggested as a measure for providing information on cardiac activity without disturbing sleep in an out-of-sleep laboratory environment [13,14,17,18,19,20,21,22,23,24]. Many previous studies validated the usefulness of ballistocardiography for overnight heartbeat monitoring [14,17,18,19,20,22,23]; however, no study has attempted to use ballistocardiographic recording for the unconstrained estimation of the ODI.…”

“…Using each type of piezoelectric sensor, 20 (20.6%), 61 (62.9%), and 16 (16.5%) recordings, respectively, were acquired with a sampling rate of 250 Hz. The load-cell-installed bed system for acquiring BCG was equipped with four-strain gauge-based load cells underneath each bed leg [12,13]. The PVDF film [14] and the EMFi sensor- based BCG acquisition systems, respectively, consisted of a 300 (length) × 300 (width) × 0.1 (height) mm PVDF film sensor and a 580 × 400 × 0.4 mm EMFi sensor.…”

A Review of Approaches for Sleep Quality Analysis

“…The arterial baroreflex, a blood pressure regulatory reflex mechanism, was analyzed by Jung et al [56] to estimate the SOL based in the observation that this mechanism has a marked influence on the HRV control with the onset of sleep. ECG and Ballistocardiogram (BCG) signals were used to estimate the HRV and the R-J interval, time interval between the ECG R peak the BCG J peak.…”

“…Gaddam et al [92] proposed the employment of four pressure sensors, placed under the legs of the bed, considering that during good sleep periods the sensors produce steady signals while the opposite happens during agitated periods, associated with poor sleep quality. Jung et al [56] used strain gauge-based load cells, placed under the bed legs, to measure BCG and infer the HRV. A device based on pressure sensors, implemented on the bed, was developed by Pino et al [93], measuring the body position and the number of respiratory cycles per minute.…”

Ballistocardiography for nonintrusive sleep structure estimation

“…So if we can detect the time when the correlation changes between heart beat interval from ECG (or from BCG) and R-J interval, we can detect the time the subject fall in sleep and estimate the sleep onset latency. First epoch whose correlation fall into negative range is detected as the first sleep epoch and estimated sleep onset latency was compared with the reference [7].…”