Does breast peritumoral tissue hold valuable information for texture analysis?
Michael Fox 1 , Peter Gibbs 1 , Martin Pickles 1 , and Lindsay W. Turnbull 1
Centre for MR Investigations, HYMS at
University of Hull, Hull, East Yorkshire, United Kingdom
The performance of GLCM texture analysis is dependent on
a well-drawn ROI segmenting a tumour from healthy
surrounding tissue. There may be changes to the
surrounding tissue in breast cancer patients which can
be an indicator of tumour status, or response
prediction. Software was developed to incrementally
expand the ROI for 100 patients and compare the
performance of GLCM texture analysis against previously
recorded values. An increase in number of differences
between TNEG patient sub-groups was found when
peritumoral tissue was included, and performance was
comparable for other patient sub-groups, suggesting
surrounding tissue should be included in analysis.
This abstract and the presentation materials are available to members only;
a login is required.