Endoscopic Robot for Distal Lung Sampling

Develop image-guided control algorithms for autonomous pulmonary sampling using a robotic endoscopic platform.
Description of the Project: 

Pulmonary infiltrates such as pus, blood, or protein, which lingers within the parenchyma of the lungs are the leading cause of pneumonia, tuberculosis.  Pulmonary infiltrates in mechanically ventilated (MV) critically ill patients in the intensive care unit (ICU) are a major diagnostic challenge and due to the poor sampling methods available. As result patients are often initiated on non-targeted therapies. Currently, there is no standarised method to diagnostically sample human lungs in mechanically ventilated critically ill patients, of which there are 20 million ventilated episodes per year.  The aim of this project is to develop a Mechatronic system for autonomous lung sampling for MV critically ill patients. The system is comprised of two interconnected physical and software components:

1) A robotic mini-bronchoscope that can autonomously adjust its shape to traverse confined spaces inside the lung airways. Our design aims to extend the reach of current technology to sample deeply nested pathologies in the distal lung. The mini-bronchoscope carries a camera, an electromagnetic tracker (EMT), and provides a working channel for “tool” delivery to the distal lung.

2) Control algorithms that employ EMT feedback and endoscopic camera images to autonomously map the inner airways and steer inside the lung to take multiple samples in a controlled and reliable manner. The system will guide to regions of the lung, use narrow channel scopes obviating any risks of airway obstruction.  

Project number: 
240011
First Supervisor: 
University: 
University of Edinburgh
First supervisor university: 
University of Edinburgh
Essential skills and knowledge: 
Robotics, Control, Strong Programming Skills. Ability to work collaboratively with clinicians.
Desirable skills and knowledge: 
Medical Robotics, Image processing, AI and Machine Vision.
References: 

Sganga J, Eng D, Graetzel C, Camarillo DB. “Deep Learning for Localization in the Lung.” (2019) arXiv:1907.08136

M. Khadem, L. Da Cruz and C. Bergeles, "Force/Velocity Manipulability Analysis for 3D Continuum Robots," 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Madrid, 2018, pp. 4920-4926.

Z. Mitros, M. Khadem, C. Seneci, S. Ourselin, L. Da Cruz and C. Bergeles, "Towards Modelling Multi-Arm Robots: Eccentric Arrangement of Concentric Tubes," 2018 7th IEEE International Conference on Biomedical Robotics and Biomechatronics (Biorob), Enschede, 2018, pp. 43-48.