Mapping and Navigation Techniques for Non-Invasive Active Endoscopic Capsules

Abstract

Yasmeen Abu-Kheil, Mapping and Navigation Techniques for Non-Invasive Active Endoscopic Capsules, Ph.D. in Engineering, Robotics Department, Khalifa University of Science and Technology, Abu Dhabi, United Arab Emirates. Capsule Endoscopy (CE) is a non-invasive procedure for gastro-intestinal (GI) diagnosis. It does not require sedation and it is comfortable and well tolerated by the patients. However, current CE practice has two main limitations: i) a high possibility of pathological miss rate due to poor image quality and the irregularity and unpredictability of the peristaltic contractions moving the capsule along the tract; ii) long time required to evaluate a huge number of images (more than 50,000) generated during a single run. Therefore, there is a need for an active capsule navigation system that would allow physicians to clearly visualize the GI tract and guide the capsule movement to enable detailed investigations of specific regions of interest. These main issues in CE could be mitigated with a tele-operated active navigation system integrating a 3D environmental reconstruction and haptic feedback that will improve the navigation process. Active navigation system will also pave the way to initiate steps for a new medical instrument that is not only controllable but also autonomous with the support of a human in the loop. In this research, a method for mapping images from a capsule-based endoscope, is proposed. The proposed technique uses visual and inertial-based data fusion to obtain a 3D map of the lumen from 2D capsule images, also paving the way for the implementation of a path planning and autonomous locomotion and inspection. Experiments on images taken from an in-vitro colon simulator were conducted. Results showed that visual and inertial-based techniques achieved promising performance in 3D mapping. The proposed technique can be applied to other active and passive capsule locomotion systems. Furthermore, a vision-based haptic control system is proposed to control the capsule movement in capsule endoscopy procedure. The vision-based system will generate 3D local maps as well as local navigation paths. The haptic guided module consists of a haptic device that allows the user to control the movement of the capsule over the generated path. The haptic guided module also helps the operator by transforming the path into a guiding virtual force. Measuring the current distance between the user input and the guiding path the haptic module will check if the user is moving away or toward the generated path and generate a feedback force. The user will also sense a repulsive virtual feedback force that keeps the user away from colon boundaries calculated from the generated local maps. Finally, the development of a virtual reality interface (CAPS-Sim) for magnetically-driven active endoscopic capsules is described. The simulator intends to provide researchers with an environment to test their design concepts as well as vision and navigation algorithms applied to endoscopic capsule procedures. CAPS-Sim is made up of four main components: (i ) haptic station, (ii ) vision station, (iii ) control station and (iv ) human machine interface (HMI). The haptic station provides the operator with i Abstract the force information from virtual or remote environments, and is also used to read the operator's hand motion. In order to provide image and force feedback, the vision station is used to capture images, generate 3D mapping and computer 'virtual interaction' forces. The capsule movements are controlled using the control station. The HMI runs a simulation engine where the capsule device and its main components are modeled and visualized. The magnetic actuation of the capsule is performed in the HMI simulation engine. The HMI also has a display unit where the colon images are visualized and rendered to the user using a 3D virtual reality glass. The endoscopic capsule image is shown on both the display unit as well as rendered to the user by oculus rift. The use of a virtual reality glass ensures the image is always in front of the doctor's eye and completely immersed into the procedure. Several scenarios of CAPS-Sim usages were preliminary reported. The current version of CAPS-Sim provides a simulator to develop and test endoscopic capsules main functions: (i ) capsule tele-operation, (ii ) haptic feedback for capsule navigation, (iii ) 3D virtual reality navigation and (iv ) region of interest (ROI) tracking. The proposed CAPS-Sim is modular and can be applied to any capsule type or any other robotic endoscopes regardless of its underlying technology.

Date of Award	Dec 2017
Original language	American English
Supervisor	Jorge Dias (Supervisor)