Research On A Wrist-type System For Detecting The Human Physiological Parameter

Hemorrhagic shock is a traumatic complication caused by severe blood loss and common in earthquake,fire,traffic accident,explosion and other disaster sites.As one of the main causes of death in trauma patients,hemorrhagic shock is characterized by rapid onset and rapid deterioration.However,the bleeding process could be halted,and patient mortality could be minimized by early identification and effective treatment.Patients will not show evident clinical symptoms in the early stages of hemorrhagic shock due to human compensating mechanism,which makes clinical diagnosis difficult for doctors.The heart rate,blood pressure,blood oxygen saturation,and other physiological indications provided by medical monitor are essential for identifying hemorrhagic shock,but medical monitors cannot be carried to disaster scenes due to their considerable size,hard application circumstances and complex field environment.This thesis reports the design of a wearable device for on-site physiological parameter monitoring of hemorrhagic shock sufferers at disaster sites,which can be used to classify injuries and improve patient survival rates.The following are some of the key points covered in the study:(1)To satisfy the injury categorization demands of patients with hemorrhagic shock at disaster sites,a physiological signal in-situ monitoring system was developed.A wristwatch with optical sensors,temperature sensors,Bluetooth low energy,microprocessor,and user interface module was created to collect temperature and pulse waves.Heart rate,blood oxygen saturation,blood pressure and core body temperature are accessible after signal processing,feature extraction and calculation model.(2)Aiming at removing the changeable pulse noise introduced by the environment,an improved sliding filter algorithm with variable span was proposed.The improved algorithm quantitatively evaluated the signal noise by the number of zeros of the differential signal,and then determined the window length and filter times of the filter.Compared with the traditional sliding filtering algorithm,the improved algorithm keeps signal details while filtering out noise in various noise situations.(3)The effect of skin individual difference,light intensity,light wavelength,measurement location on the pulse wave signal was studied.A feature-based pulse cycle screening algorithm was proposed,which reduce the burden on the signal filtering algorithm and improve the system signal processing capability.(4)A total of 11 critically patients participated in the clinical verification experiment,the results showed that the accuracy rate of heart rate was 93.55%,the accuracy of blood oxygen saturation was 87.1%,the mean bias of systolic blood pressure and diastolic blood pressure were 2 ± 7.55(mean ± SD)mmHg and 1.4 ± 7.4(mean ± SD)mmHg respectively,the accuracy of body temperature was 91%.The results demonstrated that the wristwatch for monitoring human physiological indicators has clinical benefit to some extent.