Pregnancy Labour Prediction using Magnetomyography Sensing and a Self-sorting Cybernetic Model

To date, effective means of predicting pregnancy labour continuous to lack. Magnetic field signals from uterine contractions have shown in recent studies to be able to predict labour state with a greater accuracy when compared with existing methods. This means of labour prediction methods from magnetic field signals appears to rely on a supervised learning post-processing framework whose calibration relies on an effective labelling of the training sample set. Despite its overall effectiveness, the need for sample preparation and labelling requires external intervention which in turn demands resource allocation in a clinical setting. As a potential solution to this, using a reduced electrode channel from a Magnetomyography instrumentation, we propose a multi-stage self-sorting Cybernetic model that comprises of an ensemble of various post-processing methods and underpinned by an un-supervised learning framework which allows for an automated method towards learning from the trend in the data to use to infer labour state and immanency. The results showed a comparable accuracy with that from a supervised learning method from a prior study and has produced an architecture of how an intelligent Cybernetic model can be used for labour prediction and cost saving benefits within a clinical setting.