PET image improvement using the Patch Confidence K-Nearest Neighbors Filter

Author

Sicong Yu ; Muhammed, Hamed Hamid

Author_Institution

Sch. of Technol. & Health STH, R. Inst. of Technol. KTH, Kista, Sweden

fYear

2014

fDate

1-4 June 2014

Firstpage

306

Lastpage

309

Abstract

In Positron Emission Tomography (PET), the resulted images are highly deteriorated by noise. In this study, we propose a new denoising framework using the Patch Confidence K-Nearest Neighbors Filter (PCKNN) to reduce noise in the sinogram before forwarding it to the reconstruction procedure. This method has been evaluated on a simulated PET image of a phantom, and the performance has been compared with several conventional methods in the literature. The results have shown that the PET image quality can be substantially improved in term of increased signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR).

Keywords

image denoising; image reconstruction; medical image processing; phantoms; positron emission tomography; CNR; PCKNN; PET image improvement; PET image quality; SNR; contrast-to-noise ratio; conventional methods; denoising framework; noise reduction; patch confidence K-nearest neighbors filter; phantom; positron emission tomography; reconstruction procedure; signal-to-noise ratio; sinogram; Filtering algorithms; Image reconstruction; Noise reduction; Photonics; Positron emission tomography; Signal to noise ratio;

fLanguage

English

Publisher

ieee

Conference_Titel

Biomedical and Health Informatics (BHI), 2014 IEEE-EMBS International Conference on

Conference_Location

Valencia

Type

conf

DOI

10.1109/BHI.2014.6864364

Filename

6864364