Multipoint Simultaneous Tracking of Wireless Capsule Endoscope Using Magnetic Sensor Array

Abstract

Magnetic tracking has been well studied for the localization of wireless capsule robots. However, most of the existing magnetic localization systems require a stable background magnetic field, which is not suitable for wearable applications as the movement of the human body inevitably changes the magnetic field to each sensor. The purpose of this study is to reduce the disturbance of the ambient magnetic field for localizations in different background scenarios. The proposed approach applies the magnetic density of multiple sampling points for simultaneous localization. By subtracting the magnetic field value in different positions, the background noise (BGN) field can be offset. As a result, multipoint simultaneous positioning can be achieved by using an optimization algorithm. The influence of the signal-to-noise ratio on localization accuracy has been determined through simulation analysis. Experiments first verified the feasibility of multipoint simultaneous positioning and then conducted in different geomagnetic noise and permanent magnet environments. The results show that the proposed method has been verified to be robust in different BGN environments. The proposed method is expected to be used in wearable systems for tracking magnetic capsule endoscope (MCE).