Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.771868
Title: Development of an image guidance system for laparoscopic liver surgery and evaluation of optical and computer vision techniques for the assessment of liver tissue
Author: Schneider, Crispin
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Abstract:
Introduction: Liver resection is increasingly being carried out via the laparoscopic approach (keyhole surgery) because there is mounting evidence that it benefits patients by reducing pain and length of hospitalisation. There are however ongoing concerns about oncological radicality (i.e. ability to completely remove cancer) and an inability to control massive haemorrhage. These issues can partially be attributed to a loss of sensation such as depth perception, tactile feedback and a reduced field of view. Utilisation of optical imaging and computer vision may be able to compensate for some of the lost sensory input because these modalities can facilitate visualisation of liver tissue and structural anatomy. Their use in laparoscopy is attractive because it is easy to adapt or integrate with existing technology. The aim of this thesis is to explore to what extent this technology can aid in the detection of normal and abnormal liver tissue and structures. / Methods: The current state of the art for optical imaging and computer vision in laparoscopic liver surgery is assessed in a systematic review. Evaluation of confocal laser endomicroscopy is carried out on a murine and porcine model of liver disease. Multispectral near infrared imaging is evaluated on ex-vivo liver specimen. Video magnification is assessed on a mechanical flow phantom and a porcine model of liver disease. The latter model was also employed to develop a computer vision based image guidance system for laparoscopic liver surgery. This image guidance system is further evaluated in a clinical feasibility study. Where appropriate, experimental findings are substantiated with statistical analysis. / Results: Use of confocal laser endomicroscopy enabled discrimination between cancer and normal liver tissue with a sub-millimetre precision. This technology also made it possible to verify the adequacy of thermal liver ablation. Multispectral imaging, at specific wavelengths was shown to have the potential to highlight the presence of colorectal and hepatocellular cancer. An image reprocessing algorithm is proposed to simplify visual interpretation of the resulting images. It is shown that video magnification can determine the presence of pulsatile motion but that it cannot reliably determine the extent of motion. Development and performance metrics of an image guidance system for laparoscopic liver surgery are outlined. The system was found to improve intraoperative orientation more development work is however required to enable reliable prediction of oncological margins. / Discussion: The results in this thesis indicate that confocal laser endomicroscopy and image guidance systems have reached a development stage where their intraoperative use may benefit surgeons by visualising features of liver anatomy and tissue characteristics. Video magnification and multispectral imaging require more development and suggestions are made to direct this work. It is also highlighted that it is crucial to standardise assessment methods for these technologies which will allow a more direct comparison between the outcomes of different groups. Limited imaging depth is a major restriction of these technologies but this may be overcome by combining them with preoperatively obtained imaging data. Just like laparoscopy, optical imaging and computer vision use functions of light, a shared characteristic that makes their combined use complementary.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.771868  DOI: Not available
Share: