Share Email Print

Proceedings Paper

Detection of small human cerebral cortical lesions with MRI under different levels of Gaussian smoothing: applications in epilepsy
Author(s): Diego Cantor-Rivera; Maged Goubran; Alan Kraguljac; Robert Bartha; Terry Peters
Format Member Price Non-Member Price
PDF $17.00 $21.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

The main objective of this study was to assess the effect of smoothing filter selection in Voxel-Based Morphometry studies on structural T1-weighted magnetic resonance images. Gaussian filters of 4 mm, 8 mm or 10 mm Full Width at High Maximum are commonly used, based on the assumption that the filter size should be at least twice the voxel size to obtain robust statistical results. The hypothesis of the presented work was that the selection of the smoothing filter influenced the detectability of small lesions in the brain. Mesial Temporal Sclerosis associated to Epilepsy was used as the case to demonstrate this effect. Twenty T1-weighted MRIs from the BrainWeb database were selected. A small phantom lesion was placed in the amygdala, hippocampus, or parahippocampal gyrus of ten of the images. Subsequently the images were registered to the ICBM/MNI space. After grey matter segmentation, a T-test was carried out to compare each image containing a phantom lesion with the rest of the images in the set. For each lesion the T-test was repeated with different Gaussian filter sizes. Voxel-Based Morphometry detected some of the phantom lesions. Of the three parameters considered: location,size, and intensity; it was shown that location is the dominant factor for the detection of the lesions.

Paper Details

Date Published: 12 March 2010
PDF: 10 pages
Proc. SPIE 7623, Medical Imaging 2010: Image Processing, 76233S (12 March 2010); doi: 10.1117/12.845058
Show Author Affiliations
Diego Cantor-Rivera, Robarts Research Institute (Canada)
The Univ. of Western Ontario (Canada)
Maged Goubran, Robarts Research Institute (Canada)
The Univ. of Western Ontario (Canada)
Alan Kraguljac, Robarts Research Institute (Canada)
Robert Bartha, Robarts Research Institute (Canada)
The Univ. of Western Ontario (Canada)
Terry Peters, Robarts Research Institute (Canada)
The Univ. of Western Ontario (Canada)

Published in SPIE Proceedings Vol. 7623:
Medical Imaging 2010: Image Processing
Benoit M. Dawant; David R. Haynor, Editor(s)

© SPIE. Terms of Use
Back to Top