Dynamic active constraints for robot assisted minimally invasive surgery
Author(s)
Kwok, Ka Wai
Type
Thesis
Abstract
In recent years, robot assisted Minimally Invasive Surgery (MIS) is playing an increasingly
important role in surgery. Although the benefit of reduced patient trauma and hospitalisation
with improved prognosis has been achieved through the enhanced dexterity and accuracy of
instrument manipulation by the introduction of robotic assistance, the use of current
master-slave platform has inevitably imposed the increased physical separation that deteriorates
the hand-eye coordination due to a lack of haptic feedback. To this end, the concept of Virtual
Fixtures (VFs) and Active Constraints has attracted significant research interests. It provides in
situ effective guidance of access routes to the target anatomy safely. However, its clinical
potential is only well established for procedures such as orthopaedic surgery, which are
conducted under a static frame-of-reference due to the relatively rigid anatomy involved.
The main focus of this thesis is concerned with modelling spatial constraints that are adaptive to
tissue deformation. These constraints define safe manipulation margins for an entire robot rather
than just its end-effector. An analytical framework is proposed to control an articulated flexible
robotic device. Provided with these dynamic active constraints, the framework enables the
operator to perform smooth articulation or steady navigation along curved anatomical pathways
even under rapid tissue deformation. The challenges induced by hyper-kinematic redundancy of
the robot and increased computational burden of real-time haptic rendering are addressed so that
they facilitate seamless interaction with the robot by using lower degree-of-freedom (DoF)
haptic interfacing device. Furthermore, the use of a gaze contingent paradigm is also
investigated to enhance the human-robot interaction by linking the manipulation constraints
with visual track. To demonstrate the practical nature of the proposed framework, detailed
quantitative validations were conducted on groups of subjects. Future directions and potential
improvements to the proposed techniques are finally outlined.
important role in surgery. Although the benefit of reduced patient trauma and hospitalisation
with improved prognosis has been achieved through the enhanced dexterity and accuracy of
instrument manipulation by the introduction of robotic assistance, the use of current
master-slave platform has inevitably imposed the increased physical separation that deteriorates
the hand-eye coordination due to a lack of haptic feedback. To this end, the concept of Virtual
Fixtures (VFs) and Active Constraints has attracted significant research interests. It provides in
situ effective guidance of access routes to the target anatomy safely. However, its clinical
potential is only well established for procedures such as orthopaedic surgery, which are
conducted under a static frame-of-reference due to the relatively rigid anatomy involved.
The main focus of this thesis is concerned with modelling spatial constraints that are adaptive to
tissue deformation. These constraints define safe manipulation margins for an entire robot rather
than just its end-effector. An analytical framework is proposed to control an articulated flexible
robotic device. Provided with these dynamic active constraints, the framework enables the
operator to perform smooth articulation or steady navigation along curved anatomical pathways
even under rapid tissue deformation. The challenges induced by hyper-kinematic redundancy of
the robot and increased computational burden of real-time haptic rendering are addressed so that
they facilitate seamless interaction with the robot by using lower degree-of-freedom (DoF)
haptic interfacing device. Furthermore, the use of a gaze contingent paradigm is also
investigated to enhance the human-robot interaction by linking the manipulation constraints
with visual track. To demonstrate the practical nature of the proposed framework, detailed
quantitative validations were conducted on groups of subjects. Future directions and potential
improvements to the proposed techniques are finally outlined.
Date Issued
2011-09
Date Awarded
2012-03
Copyright Statement
Attribution NoDerivatives 4.0 International Licence (CC BY-ND)
Advisor
Yang, Guang-Zhong
Elson, Dan
Publisher Department
Computing
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)