MRI simulation-based evaluation of image-processing and classification methods

被引：545

作者：

Kwan, RKS ^{[1
]}

Evans, AC ^{[1
]}

Pike, GB ^{[1
]}

机构：

[1] McGill Univ, Montreal Neurol Inst, McConnell Brain Imaging Ctr, Montreal, PQ H3A 2B4, Canada

来源：

IEEE TRANSACTIONS ON MEDICAL IMAGING | 1999年 / 18卷 / 11期

基金：

加拿大自然科学与工程研究理事会;

关键词：

image analysis; MRI; quantitative evaluation; simulation;

D O I：

10.1109/42.816072

中图分类号：

TP39 [计算机的应用];

学科分类号：

081203 ; 0835 ;

摘要：

With the increased interest in computer-aided image analysis methods, there is a greater need for objective methods of algorithm evaluation, Validation of in vivo MRI studies is complicated by a lack of reference data and the difficulty of constructing anatomically realistic physical phantoms, We present here an extensible MRI simulator that efficiently generates realistic three-dimensional (3-D) brain images using a hybrid Bloch equation and tissue template simulation that accounts for image contast, partial volume, and noise. This allows image analysis methods to be evaluated with controlled degradations of image data.

引用

页码：1085 / 1097

页数：13