Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.654445
Title: Eye tracking the interpretation of axial CT colonography
Author: Phillips , Peter William Edwin
Awarding Body: Lancaster University
Current Institution: Lancaster University
Date of Award: 2010
Availability of Full Text:
Access through EThOS:
Abstract:
Many studies of visual perception of diagnostic medical images have traditionally used modalities that are static and two-dimensional. These studies display single images, for example chest x-rays or related images grouped in a particular layout, such as mammograms. An eye tracker is used to capture eye gaze information from the observer as they view the image. Continuing advances in digital medical imaging and workstation design now permit more complex modalities to be used in diagnostic practice. In addition to standard 2D images, sectional imaging using CT or MRI can be used to compose a full 3D rendering of the subject. Image sets can also be stacked in sequence, with the observer controlling the navigation up and down the stack. Such a modality, with a discrete z-axis, can be thought of as having 2.5 dimensions. These new modalities have an interactive characteristic; searching the complete image set can no longer be performed by eye movement alone. The observers search strategy now has to include changing the data that is visible on the display at anyone time, and recalling the location of details that might not be currently visible, in addition to visual searching. This thesis addresses the issue of capturing an observer's perception of a scan as they navigated using the stack-mode visualisation. The work uses CT Colonography (CTC) as an example modality, which is very dependent on the stack mode interpretation to detect polyps in the colon. CTe is a relatively young modality, with no film-based equivalent, and relies heavily on the properties of modern eT scanner technology to generate hundreds of images for a patient volume. A key aspect of eTe interpretation is inspecting the colon wall. By tracking observer gaze, and mapping it into the scanned patient volume, this work, for the first time, calculates how much of the colon wall has been seen by an observer preforming an axial interpretation. Eye tracking in medical imaging is by no means new, but it has tended to be based around static project x-rays, or the interpretation of non-moving images. This work shows how to eye track an observer as they navigate though an axial representation of a patient volume, and how to calculate how the eye gaze path would appear in the 3D scan. This work represents the intersection of three research areas. The first is the developing field CT colonography, an imaging technique that has developed rapidly in the first decade of the 21st Century. Secondly the field of medical image perception, and the use of eye tracking to assess how a medical professional interprets a radiograph. This field has focused on film or film-equivalent imaging in its 30 year history but has not transitioned assessing multi-dimensional imaging techniques. The act of linking these two fields requires the glue -of computing technology and Human Computer Interaction (HC!), which make the third field in this multidisciplinary work. In this thesis the reader is introduced to two human body parts: the colon and the eye. An understanding of the colon and colon cancer is necessary to fully appreciate the subtleties of a particular type of imaging and interpretation. Measuring this interpretation, and indeed interpreting the interpretation, requires some familiarity with the human eye and its movements.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.654445  DOI: Not available
Share: