Use this URL to cite or link to this record in EThOS:
Title: Analysing and quantifying visual experience in medical imaging
Author: Leveque, Lucie
ISNI:       0000 0004 7962 187X
Awarding Body: Cardiff Uuniversity
Current Institution: Cardiff University
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Access from Institution:
Healthcare professionals increasingly view medical images and videos in a variety of environments. The perception and interpretation of medical visual information across all specialties, career stages, and practice settings are critical to patient care and safety. However, medical images and videos are not self-explanatory and thus need to be interpreted by humans, who are prone to errors caused by the inherent limitations of the human visual system. It is essential to understand how medical experts perceive visual content, and use this knowledge to develop new solutions to improve clinical practice. Progress has been made in the literature towards such understanding, however studies remain limited. This thesis investigates two aspects of human visual experience in medical imaging, i.e., visual quality assessment and visual attention. Visual quality assessment is important as diverse visual signal distortion may arise in medical imaging and affect the perceptual quality of visual content, and therefore potentially impact the diagnosis accuracy. We adapted existing qualitative and quantitative methods to evaluate the quality of distorted medical videos. We also analysed the impact of medical specialty on visual perception and found significant differences between specialty groups, e.g., sonographers were in general more bothered by visual distortions than radiologists. Visual attention has been studied in medical imaging using eye-tracking technology. In this thesis, we firstly investigated gaze allocation with radiologists analysing two-view mammograms and secondly assessed the impact of expertise and experience on gaze behaviour. We also evaluated to what extent state-of-the-art visual attention models can predict radiologists' gaze behaviour and showed the limitations of existing models. This thesis provides new experimental designs and statistical processes to evaluate the perception of medical images and videos, which can be used to optimise the visual experience of image readers in clinical practice.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available