Abstract #4038
Modelling diffusion-weighted MRI data from primary and metastatic ovarian tumours
Jessica M Winfield 1 , Nandita M deSouza 1,2 , Andrew N Priest 3 , Jennifer C Wakefield 2 , Charlotte Hodgkin 3 , Susan Freeman 3 , Matthew R Orton 2 , and David J Collins 1,2
1
MRI, Royal Marsden Hospital, Sutton, Surrey,
United Kingdom,
2
CRUK
Cancer Imaging Centre, Institute of Cancer Research,
Sutton, Surrey, United Kingdom,
3
Department of
Radiology, Addenbrooke's Hospital, Cambridge, United
Kingdom
Goodness-of-fit and repeatability of mono-exponential,
stretched exponential and bi-exponential models of
diffusion-weighted MRI data were assessed in 31 patients
with primary and metastatic ovarian cancer scanned
before and after chemotherapy. Coefficients of variation
of repeated baseline measurements were used to assess
repeatability of fitted parameters. Goodness-of-fit was
assessed using the Akaike Information Criterion (AIC),
which penalises additional parameters. Mono-exponential
and stretched exponential models showed good
repeatability (3 to 7 %) but the repeatability of the
bi-exponential model was poor. After excluding the
bi-exponential model, AIC analysis showed that the
stretched exponential model was preferred in pre- and
post-treatment data in all tumour sites.
This abstract and the presentation materials are available to members only;
a login is required.
Join Here
