Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.505309
Title: Hand eye coordination in surgery
Author: Leong, Jit Hung Julian John
ISNI:       0000 0004 2676 704X
Awarding Body: Imperial College London
Current Institution: Imperial College London
Date of Award: 2009
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
The coordination of the hand in response to visual target selection has always been regarded as an essential quality in a range of professional activities. This quality has thus far been elusive to objective scientific measurements, and is usually engulfed in the overall performance of the individuals. Parallels can be drawn to surgery, especially Minimally Invasive Surgery (MIS), where the physical constraints imposed by the arrangements of the instruments and visualisation methods require certain coordination skills that are unprecedented. With the current paradigm shift towards early specialisation in surgical training and shortened focused training time, selection process should identify trainees with the highest potentials in certain specific skills. Although significant effort has been made in objective assessment of surgical skills, it is only currently possible to measure surgeons' abilities at the time of assessment. It has been particularly difficult to quantify specific details of hand-eye coordination and assess innate ability of future skills development. The purpose of this thesis is to examine hand-eye coordination in laboratory-based simulations, with a particular emphasis on details that are important to MIS. In order to understand the challenges of visuomotor coordination, movement trajectory errors have been used to provide an insight into the innate coordinate mapping of the brain. In MIS, novel spatial transformations, due to a combination of distorted endoscopic image projections and the "fulcrum" effect of the instruments, accentuate movement generation errors. Obvious differences in the quality of movement trajectories have been observed between novices and experts in MIS, however, this is difficult to measure quantitatively. A Hidden Markov Model (HMM) is used in this thesis to reveal the underlying characteristic movement details of a particular MIS manoeuvre and how such features are exaggerated by the introduction of rotation in the endoscopic camera. The proposed method has demonstrated the feasibility of measuring movement trajectory quality by machine learning techniques without prior arbitrary classification of expertise. Experimental results have highlighted these changes in novice laparoscopic surgeons, even after a short period of training. The intricate relationship between the hands and the eyes changes when learning a skilled visuomotor task has been previously studied. Reactive eye movement, when visual input is used primarily as a feedback mechanism for error correction, implies difficulties in hand-eye coordination. As the brain learns to adapt to this new coordinate map, eye movements then become predictive of the action generated. The concept of measuring this spatiotemporal relationship is introduced as a measure of hand-eye coordination in MIS, by comparing the Target Distance Function (TDF) between the eye fixation and the instrument tip position on the laparoscopic screen. Further validation of this concept using high fidelity experimental tasks is presented, where higher cognitive influence and multiple target selection increase the complexity of the data analysis. To this end, Granger-causality is presented as a measure of the predictability of the instrument movement with the eye fixation pattern. Partial Directed Coherence (PDC), a frequency-domain variation of Granger-causality, is used for the first time to measure hand-eye coordination. Experimental results are used to establish the strengths and potential pitfalls of the technique. To further enhance the accuracy of this measurement, a modified Jensen-Shannon Divergence (JSD) measure has been developed for enhancing the signal matching algorithm and trajectory segmentations. The proposed framework incorporates high frequency noise filtering, which represents non-purposeful hand and eye movements. The accuracy of the technique has been demonstrated by quantitative measurement of multiple laparoscopic tasks by expert and novice surgeons. Experimental results supporting visual search behavioural theory are presented, as this underpins the target selection process immediately prior to visual motor action generation. The effects of specialisation and experience on visual search patterns are also examined. Finally, pilot results from functional brain imaging are presented, where the Posterior Parietal Cortical (PPC) activation is measured using optical spectroscopy techniques. PPC has been demonstrated to involve in the calculation of the coordinate transformations between the visual and motor systems, which establishes the possibilities of exciting future studies in hand-eye coordination.
Supervisor: Darzi, Ara ; Yang, Guang-Zhong ; Emery, Roger Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.505309  DOI: Not available
Share: