Across the world, healthcare costs are projected to continue to increase, and the pressure on the healthcare system is only going to grow in intensity as the rate of growth of elderly population increases in the coming decades. As an example, when people age one possible condition that they may experience is sleep-disordered breathing (SDB). SDB, better known as the obstructive sleep apnea (OSA) syndrome, and associated cardiovascular complications are among the most common clinical disorders. The gold-standard approach to accurately diagnose OSA, is polysomnography (PSG), a test that should be performed in a specialist sleep clinic and requires a complete overnight stay at the clinic. The PSG system can provide accurate and real-time data; however, it introduces several challenges such as complexity, invasiveness, excessive cost, and absence of privacy. Technological advancements in hardware and software enable noninvasive and unobtrusive sensing of vital signs. An alternative approach which may help diagnose OSA and other cardiovascular diseases is the ballistocardiography. The ballistocardiogram (BCG) signal captures the ballistic forces of the heart caused by the sudden ejection of blood into the great vessels with each heartbeat, breathing, and body movement. In recent years, BCG sensors such as polyvinylidene fluoride film-based sensors, electromechanical films, strain Gauges, hydraulic sensors, microbend fiber-optic sensors as well as fiber Bragg grating sensors have been integrated within ambient locations such as mattresses, pillows, chairs, beds, or even weighing scales, to capture BCG signals, and thereby measure vital signs. Analysis of the BCG signal is a challenging process, despite being a more convenient and comfortable method of vital signs monitoring. In practice, BCG sensors are placed under bed mattresses for sleep tracking, and hence several factors, e.g., mattress thickness, body movements, motion artifacts, bed-partners, etc. can deteriorate the signal. In this paper, we introduce the sensors that are being used for obtaining BCG signals. We also present an in-depth review of the signal processing methods as applied to the various sensors, to analyze the BCG signal and extract physiological parameters such heart rate and breathing rate, as well as determining sleep stages. Besides, we recommend which methods are more suitable for processing BCG signals due to their nonlinear and nonstationary characteristics.
Read More on PubMed