Politecnico di Torino (logo)

Design and development of a real-time localisation algorithm for the magnetic manipulation of the Magnetic Flexible Endoscope.

Luca Audino

Design and development of a real-time localisation algorithm for the magnetic manipulation of the Magnetic Flexible Endoscope.

Rel. Alessandro Rizzo. Politecnico di Torino, Corso di laurea magistrale in Mechatronic Engineering (Ingegneria Meccatronica), 2020

PDF (Tesi_di_laurea) - Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives.

Download (18MB) | Preview

Colorectal cancer (CRC) is the third leading cause of cancer death worldwide, and its incidence is constantly rising in developing nations. It usually starts as a benign tumour in the form of a gastrointestinal polyp, which becomes cancerous over time. Colonoscopy is the most common recommended screening procedure. During this exam, a semi-rigid endoscope with diagnostic and therapeutic capabilities is inserted into the patient via the rectum and pushed forward through the large intestine by the physician. However, owing to the method of actuation, current endoscopes often cause tissue damages and patient discomfort. In the past twenty years, a wide range of tethered and wireless devices have been developed to mitigate these limitations, so to decrease the risk associated with the colonoscopy. In this area of research, magnetically actuated and controlled mesoscale capsules have shown to have the potential to revolutionize GI endoscopy and transform the perception of patients toward this recommended screening procedure. Among these devices, the Magnetic Flexible Endoscope (MFE), designed and developed by the Science and Technologies Of Robotics in Medicine (STORM) Lab UK and USA, uses a purpose-built magnetic system mounted on the end-effector of a robotic manipulator to control a tethered capsule endoscope. To apply the necessary forces and torques to the endoscopic tip and to enable a precise polyps’ detection, the accurate real-time pose estimation of the device is crucial. Although the currently implemented localisation algorithm solves important drawbacks faced by its previous versions (Singularity regions of the magnetic field and attitude initialization), it requires high computational effort because of the Particle Filter based approach; This last feature represents a constraint for real-time performance. The work presented in this Master Thesis is focused on the design of a new localisation algorithm for the MFE based on Kalman Filters. Due to the multiple implementations of these estimators, a preliminary research is carried out to identify the most suitable version according to the nonlinear mathematical models used for the prediction of the capsule pose. Special attention is given to the Kalman Filter (KF), Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF). This prior study highlights the impossibility to use traditional EKF and UKF with the currently used magnetic field models and suggests a feasible algorithm based on a Kalman Filter. The estimation method developed hereby represents a proof of concept for the application of this widely used estimator to achieve robust and real-time performances with a lower computational effort. Although static and dynamic tests have been performed to verify the correctness of the algorithm, this localisation method needs to be optimized on the robotic platform. In fact, due to the containment measures against the spread of the new coronavirus COVID-19, the physical Magnetic Flexible Endoscope has not been employed in the project. To overcome this limitation, a detailed ROS simulator of the MFE has been developed in Python and used for the design of the new localisation algorithm. The main target of the simulator, which involves the use of an IMU module and an STM32 Nucleo board, is to generate at 100Hz a set of inertial and magnetic field data as similar as possible to the ones retrieved by the sensors embedded in the physical capsule.

Relators: Alessandro Rizzo
Academic year: 2020/21
Publication type: Electronic
Number of Pages: 98
Corso di laurea: Corso di laurea magistrale in Mechatronic Engineering (Ingegneria Meccatronica)
Classe di laurea: New organization > Master science > LM-25 - AUTOMATION ENGINEERING
Ente in cotutela: University of Leeds (REGNO UNITO)
Aziende collaboratrici: University of Leeds
URI: http://webthesis.biblio.polito.it/id/eprint/16050
Modify record (reserved for operators) Modify record (reserved for operators)