Computer vision methods for improved detection and characterisation of cancer in breast MRI
Speaker: Andrew Hill, ITEE
When: 2007-07-04 10:00:00
Venue: 78-420
Host: Prof Stuart Crozier
Abstract:This PhD research constitutes part of a larger research project,
known as the Breast MRI project, within the Electromagnetics and
Imaging Research Division within the School of ITEE. The aim of the
project is to improve the specificity and the sensitivity of breast
MRI, and therefore its clinical utility, through the use of computer
vision (CV) -- computerised image analysis and pattern recognition
-- techniques and the integration of information concerning breast
tissue morphology and contrast-enhancement kinetics. Specifically
this project will aim to increase the selectivity of existing lesion
segmentation approaches by examining the areas of motion correction,
segmentation of breast tissue, pharmaco-kinetic modeling, and bias
field correction.
To date this research has consisted of the implementation of two
novel registration techniques and the publication of a framework to
quantitatively compare the accuracy of different registrations. The
second segment of work, the segmentation of the breast volume, will
help reduce false positives within automated analysis by removing
several sources of highlighting tissue, such as the liver and
auxcipital nodes, from the analysed image volume. The third segment
of work, pharmaco-kinetic modeling has two aims. Firstly to evaluate
the various pre-existing models, and secondly to provide a means of
estimating the patient dependent component of the global parameters
within the Tofts model. Finally means of correcting the bias field
will be examined as in many situations direct comparison between
different segments of the image can be hindered by the induced bias
effects.
Biography:(biography unavailable)
Type: Ph.D confirmation
Contact:Prof Stuart Crozier, seminar host (stuart@itee.uq.edu.au)
or Guido Governatori (ITEE seminar co-ordinator)
(guido@itee.uq.edu.au)