Key Technology Development Of Smart Mattress Based On Fiber Optic Micro-pressure Sensing

With the development of the internet and technology,the transformation and innovation of medical industry are promoted and then “medical wisdom” has emerged.The key to achieve “medical wisdom” lies in the development of household and convenient medical monitoring equipment,which allows physiological parameters to be shown intuitively for diagnosing the health state of the subjects in time.Electrical sensors have further developed in health monitoring field.However,electrical sensors suffer from electromagnetic interference,short life,and easy-corrosion,which is not suitable for long-term medical monitoring.Fiber optic sensors have the advantages of flexibility,small,light,electromagnetic interference immunity and high resistance to moisture,which have been widely applied in health monitoring field.Fiber optic micropressure sensors is sensitive to the weak pressure,strain,and vibration induced by physiological activities,which have become the first choice of developing non-invasive vital signs monitoring equipment.Besides,the present vital signs extraction algorithms are based on electrical sensing system,and have the drawbacks of imperfect function and difficult realization.So,these algorithms are not suitable for integration into fiber optic vital signs monitoring system.In view of above requirement of health monitoring,we have proposed a smart mattress system based on fiber optic Mach-zehnder interferometer.Based on the matched Multi-parameter feature extraction algorithm,the on-line simultaneous monitoring of heartbeat and respiration indicators has been achieved.The mattress is promising in the early detection and prevention of cardiac and respiratory diseases.The main contents of this thesis are listed as follows:Firstly,the working principle of fiber optic micro-pressure sensor has been studied.Specifically,the photo-elastic effect,sensing principle of fiber optic Mach-zehnder interferometer,and phase demodulation method have been respectively studied.The theoretical model of fiber packaging based on highly elastic polymer has been built.Through the model and simulation achieved by Matlab software,it is proved that the sensitization effect of PVC material packaging has reached 90 times.Secondly,a feature extraction algorithm is designed for obtaining four indicators,which includes the four software modules of phase demodulation,activity states monitoring,waveform recovering,and feature indicator extraction.The specific function of the four software modules is as follow.In the module of phase demodulation,the raw phase is demodulated by adopting the 3×3 coupler and the differentiate and cross-multiplying demodulation algorithm.In the module of activity states monitoring,the four activity states of nobody,on bed,body movement,and off bed are judged by utilizing high-pass filter and calculating the frequency spectrum energy of the mixed signal.In the module of waveform recovering,the respiration and heartbeat waveforms are recovered through median fitting and difference operation.In the module of feature indicator extraction,the four indicators of respiration rate,heartbeat rate,respiration amplitude and heartbeat amplitude are extracted by finding peaks in frequency domain and time domain,respectively.Thirdly,we have designed the system structure of fiber optic mattress and the structure of system software.The mattress system consists of a sensing mattress and a terminal box,in which the sensing mattress is fabricated by packaging sensitization structure based on PVC material and silica gel material,and the terminal box includes an optical demodulation subsystem and a signal processing module.The system software includes feature extraction algorithm module,current signs display module and recorded signs display module.Specifically,the feature extraction algorithm module calls the feature extraction algorithm based on Labview software.Developed by utilizing Labview software,the user interface,constructed by the current signs display module and recorded signs display module,includes a current vital signs interface and a record interface,which has the functions of displaying current vital signs,the alarm of abnormal conditions,and recording historical vital signs.Finally,the clinic experiments have been carried out in terms of the functions and performance of the mattress system.By monitoring the four activity states of nobody,on bed,body movement,and off bed,the ability of the activity states monitoring algorithm in judging the four activity states are proved.Meanwhile,the feature extraction algorithm is capable of diagnosing the symptoms of bradycardia,tachycardia,polypnea and apnoea by analyzing four indicators related to heartbeat and respiration.Further,the statistic analysis has been applied to the measuring results of eighteen subjects.The results shows that the max errors of respiration rate and heart rate are respectively 2 bpm and 1 bpm and the largest standard deviations of respiration rate and heart rate are respectively 2.6 bpm and 3.0 bpm,which is acceptable according to the American National Standard ANSI/AAMI.The Bland-Altman analysis method has been also applied to the measuring results.The results indicates that 100% of values lies in the limits of agreement of HR and RR,which verifies that the mattress has good consistency,compared with the reference measurement.Moreover,two experiments have been carried out,including the texting of long-term continuous monitoring and the response consistency testing at different positions of the mattress.Experimental results indicate that the difference of heart rate and respiration rate measured at six positions of mattress is less than 5% and the system has good consistency response in measuring heart rate and respiration rate.