|
Activity Number:
|
117
|
|
Type:
|
Invited
|
|
Date/Time:
|
Monday, July 30, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Committee on Applied Statisticians
|
| Abstract - #307803 |
|
Title:
|
Some Statistical Issues with DCE-MRI Reproducibility Studies in Oncology
|
|
Author(s):
|
William L. Mietlowski*+ and Yuhui Ma and Theodore C. Pellas and Ching-Ray Yu
|
|
Companies:
|
Novartis Pharmaceuticals and University of Medicine and Dentistry of New Jersey and Novartis Pharmaceuticals and Rutgers University
|
|
Address:
|
1 Health Plaza, East Hanover, NJ, 07936,
|
|
Keywords:
|
DCE-MRI ; reproducibility ; repeatability ; distribution-free ; tolerance intervals
|
|
Abstract:
|
Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) has been used to a tool to evaluate the effect of anti-angiogenic therapies in cancer. A workshop report proposing standards for DCE-MRI (Leach et al 2005 British J Cancer) recommends incorporation of two baselines (DCE-MRI reproducibility studies) to estimate individual patient reproducibility. This degree of reproducibility is often used to determine the degree of change necessary for a patient to be classified as a DCE-MRI responder. Two methods have been used to estimate individual patient reproducibility in the imaging literature. Galbraith et al (2002 NMR Biomedicine) proposed a procedure based on repeatability while Evelhoch et al (2004 Clinical Cancer Research) proposed using the intra-patient CV. Statistical issues with these approaches will be discussed and compared with a distribution-free tolerance limit approach.
|
