The Preliminary Research On Monitoring Technology Of The Non-contact Sleep State Based On Bioradar

Sleep is indispensable to life which accounts for a third of life. In the state of sleepthe body’s physiological and psychological activities are in resting state which providesthe condition for the analysis of human sleep. Through the analysis physiological andpsychological signal of human sleep state can find some disease problems or pathogenicfactors which are hard to notice to provide some valuable information for the diagnosis ofthe related disease.The incidence of sleep disorders is increasing makes people aware of the diseases.Sleep apnea syndrome is one of them. Sleep apnea is the phenomenon that the number ofoccurrences of apnea is more than5times per hour, or that occur more than30timesduring7h sleep. If this phenomenon occurs frequently, it would cause serious consequences to human health.The gold method used in monitoring sleep is polysomnography (PSG). Sleepmedicine has relied heavily on laboratory-based PSG as a primary research tool. However,it needs multiple electrodes or wires to constrain the subject for a long time, and themonitoring often takes place in the sleep laboratory and does not apply to family care. Therelatively simple monitoring method is using the actigraphy, but it has some limitations onto get the data. In order to monitor human sleep in more comfortable, natural environment,to monitoring human sleep, the type of mattress polysomnography system non-electrodehas been developed, but this method requires special beds or mattresses which lead to theargument that the difference in feeling of beds affects the patient sleep. Hence, therecording may not be a true indicator of the patients "daily" sleep. Biological radar canpenetrate non-metallic medium (such as clothing, bedding, etc) to detect humanrespiratory signal without electrodes or wires to contact the subject, and has somesuperiority in terms of family sleep monitoring. Therefore, we use micro power biologicalradar to monitor human sleep in this paper, and expect to achieve the goal of preliminaryscreening for sleep apnea. At the same time, due to sleep apnea happened under thecondition of the human don’t even know it, it could threaten to the life safety, if thebreathing can’t restore for a long time, so, the alarm system is required at this time, towake up the human body from sleep in time. We finished the following work in this study.Ⅰ Established the synchronous acquisition system, to analyze the reliability of therespiration signal which detected by biological radar.Established biological radar and respiration bind synchronous signal acquisitionsystem. This system respectively detected signals from four postures of human body insleep, we did the respiratory signal analysis spectrum and correlation analysis between thetwo signals in this study. The result shows that the signal collected by the radar containsabundant breathing component, and has strong correlation with the signal collected byrespiration bind. This also shows that the respiratory signal collected by the radar has highreliability, which lay the foundation for subsequent work.Ⅱ Adopted the method of the energy spectrum to distinguish sleep apnea. In the time domain, the signal energy has the additivity, but it vulnerable to theinfluence of external disturbance, so, before using the energy spectrum method, we shouldfirst roughly verify the range of signal-to-noise ratio (SNR), in which the method isfeasible. In this experiment, we got the range of SNR by using the energy spectrummethod to determine structure respiratory signal, experience certificate that this methodhas a wide applicable range of SNR, and the SNR of the signals detected by radar all arein this range. So, we can use this method to analysis the signal collected by the radar. Theresult shows that the method is suitable for preliminary judgment of apnea.Ⅲ Adopted the method of the maximum frequency to distinguish movements.Human body tend to adjust the sleep posture, or take a deep breath to ease of thedifficulties in respiration，the times of movements occurred in sleep will affect the state ofspirit. So, the study on the times of movements can help to judge the state of spirit nextday, thus infer that the subjects are suitable to do some highly focused work, such asdriving or flying. The frequency of the signals is different between the normal respirationand the movement, so, we analyze the signals in the frequency domain. We firstly find outthe maximum frequency of each signal in this study, and woke out the times of respiration,then compare the times we got with the timed of normal respiration, thus infer thatwhether the movement occurred.Ⅳ Adopted the method of the information entropy to analyze reliability of theprocessing results by the energy spectrum method or the maximum frequency method.We use the characteristics of information entropy that it can determine the amount ofinformation contained in the signals, to detect the apnea signal or the movement signal.The information entropy can determine the beginning and end of apnea or movementsignal, and analyze the processing results with the results by the energy spectrum methodor the maximum frequency method processed, then we got that the processing results bythe energy spectrum method or the maximum frequency method processed is reliable. Theresult shows that these two methods are suitable for judgment of apnea and movement.Ⅴ Used energy spectrum method to process multiple sets of apnea data (differentsubjects). The result shows that it has higher accuracy by this method process, at the same time it shows that biological radar can be used to monitor sleep, and can carry on thepreliminary judgment for apnea.There are the main innovation points in this paper.Ⅰ Adopt the method of the energy spectrum to distinguish sleep apnea base on thebiological radar monitoring.Ⅱ Adopt the method of the maximum frequency to distinguish movement base onthe biological radar monitoring, put forward the corresponding algorithm.Ⅲ Put forward that adopt the energy spectrum to analyze reliability of theprocessing results by the energy spectrum method or the maximum frequency method.