Computer controlled redundant endoscopy

Abstract

The fundamental limitation of current flexible endoscopes is the inability to know the position and orientation of all parts of the device. The two main consequences are, first, that the view from the end of the scope must be interpreted both for depth and orientation with the corresponding implication for trying to move the tip, and second, that the position of the flexible shaft behind the tip is unknown, with the associated possibility of collateral damage. These weaknesses can be considered as statements of the lack of understanding of the kinematics of the device. The path which the endoscope tip should take is not clearly defined and the ability to control the device to follow a unique path does not exist. This paper proposes a solution that would bring endoscopes into the set of devices that can be examined and controlled using standard robotic tools. However, an endoscope is best categorised as a highly redundant manipulator, with flexible links. The stable solution to the path planning of highly redundant mechanisms is itself a partially unsolved problem, although the use of potential field techniques linked with the surgical guidance modalities of magnetic resonance imaging and computer aided tomography indicates that a computer controlled endoscope is feasible.