Implementation of statistical clustering for voxel-based kinetic analysis in PET

Author

Kimura, Yuichi ; Noshi, Yasuhiro ; Ishii, Kenji ; Oda, Keiichi ; Ishiwata, Kiichi

Author_Institution

PET Center, Tokyo Metropolitan Inst. of Gerontology, Japan

Volume

2

fYear

2002

fDate

10-16 Nov. 2002

Firstpage

1322

Abstract

Voxel-based kinetic analysis to a dynamic data derived from PET has a potential capability to visualize some functionality of living tissue. But bad noisy statistic in voxel-vased PET data and a huge number of calculation time make it impractical to apply an ordinal ROI-based kinetic analysis to voxel-by-voxel. The novel method for voxel-based kinetic analysis is proposed, clustering analysys for kinetics (CAKS). In CAKS, voxels are clustered by its time history, and parameter estimation is applied in this cluster. As the result, CAKS can derive physiologically acceptable parametric images in 20 minutes. We can conclude that CAKS is a practical method to realize voxel-based kinetic analysis.

Keywords

medical image processing; parameter estimation; positron emission tomography; PET; living tissue; parameter estimation; parametric image; statistical clustering; time history; voxel-based kinetic analysis; Clustering algorithms; Helium; Image analysis; Kinetic theory; Noise measurement; Parameter estimation; Plasma measurements; Pollution measurement; Positron emission tomography; Statistics;

fLanguage

English

Publisher

ieee

Conference_Titel

Nuclear Science Symposium Conference Record, 2002 IEEE

Print_ISBN

0-7803-7636-6

Type

conf

DOI

10.1109/NSSMIC.2002.1239562

Filename

1239562