Preliminary study to detect fetal movement by using acceleration sensor and MEMS microphone

Abstract

Fetal movement is an indicator of well-being. If the fetus does not move for a certain number, it may indicate a pregnancy risk. Acceleration measurement is a popular technique for detecting the vibration due to the movement. However, the comparison to use acceleration with other techniques e.g. fetal-movement sound has not been performed to date. In this study, acceleration sensors and MEMS microphones were used to detect three actions performed on the subject's abdomen: flicking, tapping, and knocking with finger(s), and two activities without performing these actions: subject speaking and walking. These actions and activities were carried on a female healthy subject in order to simulate the actions due to fetal movement. The maximum accuracy of the use of acceleration sensors was 69.96% for the tapping action; while that of MEMS microphones was 71.11% for the flicking action. The very poor accuracy was found for the knocking action with using MEMS microphones, i.e. 31.11%. Using the MEMS microphones was very robust to the interference from speaking and walking, while using the acceleration sensors was very susceptible to walking. Therefore, to develop a device for counting fetal movement, employing both sensors would be recommended in order to sense various types of movement. © 2018 IEEE