Adaptive Wireless Biomedical Capsule Localization and Tracking

Abstract

Wireless capsule endoscopy systems have been shown as a gold step to develop future wireless biomedical multitask robotic capsules, which will be utilized in micro surgery, drug delivery, biopsy and multitasks of the endoscopy. In such wireless capsule endoscopy systems, one of the most challenging problems is accurate localization and tracking of the capsule inside the human body. In this thesis, we focus on robotic biomedical capsule localization and tracking using range measurements via electromagetic wave and magnetic strength based sensors. First, a literature review of existing localization techniques with their merits and limitations is presented. Then, a novel geometric environmental coefficient estimation technique is introduced for time of flight (TOF) and received signal strength (RSS) based range measurement. Utilizing the proposed environmental coefficient estimation technique, a 3D wireless biomedical capsule localization and tracking scheme is designed based on a discrete adaptive recursive least square algorithm with forgetting factor. The comparison between localization with novel coefficient estimation technique and localization with known coefficient is provided to demonstrate the proposed technique’s efficiency. Later, as an alternative to TOF and RSS based sensors, use of magnetic strength based sensors is considered. We analyze and demonstrate the performance of the proposed techniques and designs in various scenarios simulated in Matlab/Simulink environment.