MRI - STABILITY OF 3 SUPERVISED SEGMENTATION TECHNIQUES

被引：113

作者：

CLARKE, LP

VELTHUIZEN, RP

PHUPHANICH, S

SCHELLENBERG, JD

ARRINGTON, JA

SILBIGER, M

机构：

[1] Center for Engineering and Medical Image Analysis (CEMIA) of the Colleges of Engineering and Medicin, University of South Florida, Tampa

[2] H. Lee Moffitt Cancer Center, Research Institute, Department of Neurology, Tampa

来源：

MAGNETIC RESONANCE IMAGING | 1993年 / 11卷 / 01期

关键词：

MRI; IMAGE SEGMENTATION; PATTERN RECOGNITION METHODS; NEURAL NETWORKS;

D O I：

10.1016/0730-725X(93)90417-C

中图分类号：

R8 [特种医学]; R445 [影像诊断学];

学科分类号：

1002 ; 100207 ; 1009 ;

摘要：

Supervised segmentation methods from three families of pattern recognition techniques were used to segment multispectral MRI data. Studied were the maximum likelihood method (MLM), k-nearest neighbors (k-NN), and a back-propagation artificial neural net (ANN). Performance was measured in terms of execution speed, and stability for the selection of training data, namely, region of interest (ROI) selection, and interslice and interpatient classifications. MLM proved to have the smallest execution times, but demonstrated the least stability. k-NN showed the best stability for training data selection. To evaluste the segmentation techniques, multispectral images were used of normal volunteers and patients with gliomas, the latter with and without MR contrast material. All measures applied indicated that k-NN provides the best results.

引用

页码：95 / 106

页数：12

共 34 条

[1]

Hyman, Kurland, Levy, Shoop, Characterization of normal brain tissue using seven calculated MRI parameters and a statistical analysis system, Magn. Reson. Med., 11, pp. 22-34, (1989)

[2]

Rajanayagam, Fabry, Gore, In-vivo quantitation of water content in muscle tissue by NMR imaging, Magn. Reson. Imaging, 9, pp. 621-625, (1991)

[3]

Kohn, Tanna, Herman, Resnick, Mozley, Gur, Alavi, Zimmerman, Gur, Analysis of brain and cerebrospinal fluid volumes with MR imaging, Radiology, 178, 1, pp. 115-122, (1991)

[4]

O'Donnell, Gore, Adams, Toward and automated analysis system for nuclear magnetic resonance imaging. II. Initial segmentation algorithm, Med. Phys., 13, 3, pp. 293-297, (1986)

[5]

Schellenberg, Naylor, Clarke, Application of artificial neural networks for tissue classification from multispectral magnetic resonance images of the head, Proceedings of the Third Annual IEEE Symposium on Computer-Based Medical Systems, (1990)

[6]

Clarke, Comparison of Bayesian maximum likelihood (MLM) and artificial neural network for supervised tissue classification and 3D segmentation by MRI, 33rd Annual Scientific Symposium of American Association of Physicists in Medicine (AAPM), 18, (1991)

[7]

Vannier, Butterfield, Jordan, Murphy, Levitt, Gado, Multispectral analysis of magnetic resonance images, Radiology, 154, pp. 221-224, (1985)

[8]

Bensaid, Hall, Clarke, Velthuizen, MRI Segmentation using supervised and unsupervised methods, Proceedings of the 13th Annual IEEE Eng Med & Biol., (1991)

[9]

Levin, Hu, Tan, Galhotra, Pelizzari, The brain: integrated three-dimensional display of MR and PET images, Radiology, 172, pp. 783-789, (1989)

[10]

Xiaoping, Alperin, Levin, Visualization of MR angiographic data with segmentation and volume rendering techniques, Journal of Magnetic Resonance Imaging, 1, 5, (1991)

← 1 2 3 4 →