Detecting Fetal Movements Using Non-Invasive Accelerometers: A Preliminary Analysis

Abstract

Monitoring fetal movement is important to assess fetal health. Standard clinical fetal monitoring technologies include ultrasound imaging and cardiotocography. Both have limited prognostic value and require significant health resources. We have recently developed a low-cost, passive, non-invasive system to monitor fetal activity, and therefore fetal health. This accelerometer-based system does not require trained operators and can be used outside a clinic. This work is a preliminary study to develop a method to automatically detect fetal movement using this new accelerometer system. We assess the efficacy of using a threshold method over a range of different frequency bands. We also examine using a set of statistical features for a detection method. Our results indicate that neither method performs sufficiently well to automatically detect fetal movement.