Activity Number:
|
224
|
Type:
|
Invited
|
Date/Time:
|
Monday, August 1, 2016 : 2:00 PM to 3:50 PM
|
Sponsor:
|
Section on Statistics in Imaging
|
Abstract #318166
|
View Presentation
|
Title:
|
Statistics on Tumor Images Using Trees and Shapes of Parameterized Curves
|
Author(s):
|
Karthik Bharath* and Veera Baladandayuthapani and Arvind Rao and Sebastian A. Kurtek and Dipak Dey
|
Companies:
|
University of Nottingham and MD Anderson Cancer Center and MD Anderson Cancer Center and The Ohio State University and University of Connecticut
|
Keywords:
|
radiomics ;
tumour shapes ;
tree-structured data ;
MRI
|
Abstract:
|
We present techniques to model and perform inference on MR images of tumours based on tree-based representations of voxel intensities, and parameterized curves for tumour shapes. Such holistic representations of images, in contrast to low-dimensional numerical summaries, capture several interesting features of images. Statistical procedures on the non-Euclidean spaces of trees and curves are described in the context of images of brain tumours obtained from patients in the presence of clinical and genetic covariates.
|
Authors who are presenting talks have a * after their name.