Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.756085
Title: Visual tracking of instruments in minimally invasive surgery
Author: Allan, M. H.
ISNI:       0000 0004 7429 0418
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2017
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Abstract:
Reducing access trauma has been a focal point for modern surgery and tackling the challenges that arise from new operating techniques and instruments is an exciting and open area of research. Lack of awareness and control from indirect manipulation and visualization has created a need to augment the surgeon's understanding and perception of how their instruments interact with the patient's anatomy but current methods of achieving this are inaccurate and difficult to integrate into the surgical workflow. Visual methods have the potential to recover the position and orientation of the instruments directly in the reference frame of the observing camera without the need to introduce additional hardware to the operating room and perform complex calibration steps. This thesis explores how this problem can be solved with the fusion of coarse region and fine scale point features to enable the recovery of both the rigid and articulated degrees of freedom of laparoscopic and robotic instruments using only images provided by the surgical camera. Extensive experiments on different image features are used to determine suitable representations for reliable and robust pose estimation. Using this information a novel framework is presented which estimates 3D pose with a region matching scheme while using frame-to-frame optical flow to account for challenges due to symmetry in the instrument design. The kinematic structure of articulated robotic instruments is also used to track the movement of the head and claspers. The robustness of this method was evaluated on calibrated ex-vivo images and in-vivo sequences and comparative studies are performed with state-of-the-art kinematic assisted tracking methods.
Supervisor: Stoyanov, D. ; Ourselin, S. ; Hawkes, D. ; Kelly, J. Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.756085  DOI: Not available
Share: