Design And Research Of Fetal Movements Signal Acquisition And Preprocessing System

Fetal movement is an important vital sign of the fetus that has formed in the early stages of the fetus and continues throughout the pregnancy.Abnormality of fetal movement is often a signal that the fetus is in danger,early detection of abnormal fetal movement can effectively reduce the lethality of the fetus.Currently,there are mainly two methods for detecting fetal movement.One is the count of mothers based on their own perceptions,this method is relatively simple to implement,but the counting of fetal movement is very inaccurate;The other is Doppler imaging technology which is commonly used in hospital obstetric,requires the participation of physicians,and can not do long-term monitoring,the cost of it is more expensive.This paper presents a new fetal movement signal acquisition program.In this program,the multi-axis accelerometer is sewed in the fabric,close to the abdomen of the pregnant woman,collecting the signal transmitted by the fetal activity to the abdominal wall.The program is wearable and allows for long-term monitoring without the assistance of a physician.In order to achieve this program,this paper mainly studies the following aspects:1)For the signal acquisition system design,this article will be described in two aspects,inclcuding the hardware structure and software implementation.The controller is Freescale’s Kinetis KL16 microcontroller which acquire fetal movement signals from the abdominal wall of pregnant women via 4 accelerometers,packetizes the data,and then controls Bluetooth to send the data to the mobile device.By designing the data packet protocol for communication between the microcontroller and Bluetooth and improving the Bluetooth protocol stack,accuracy and real-time of fetal movement data is ensured in the transmission process.2)Design the experiment for applying fetal movement equipment,recruit volunteers for clinical signal acquisition and preprocess the collected data.Design suitable preprocess methods to remove some easily separated noise and maternal interference signals.The preprocessed data is divided into a block of time segments,the fetal movement can be captured with the Doppler imaging equipment records.3)Research common non-stationary signal analysis methods including time-frequency analysis and signal sparse decomposition,find out the signal characteristics of fetal movement and maternalpseudo-signal characteristics.According to the characteristics of fetal movement signal and pseudo-movement signal,a fetal movement identification methodwas designed,and the recognition rate of fetal movement would be evaluated.The number of fetal movement collected by Doppler was used as the evaluation criteria.The final assessment showed that fetal movement recognition rate ofthe method is much higher than that of maternal perception.