Study on useful and high accuracy noise reduction method in SPECT and PET

2018 Fiscal Year Final Research Report

• Project/Area Number: 16K19865
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Radiation science
• Research Institution: Kyoto College of Medical Science
• Principal Investigator: MATSUMOTO Keiichi 京都医療科学大学, 医療科学部, 准教授 (60393344)
• Research Collaborator: NAKAMURA Hitomi ; OGINO Yukako
• Project Period (FY): 2016-04-01 – 2019-03-31
• Keywords: 量子雑音除去 / ウェーブレット変換 / 核医学画像 / SPECT / PET

Outline of Final Research Achievements

The purpose of this study was to develop a novel Poisson noise reduction method of the nuclear medicine images based Bayes Wavelet Shrink technique. A present method was evaluated to a digital brain phantom (produced by Japanese Society of Nuclear Medicine Working Group) with statistical noise and a Hoffman 3-D brain phantom. Quality of the reconstructed images was evaluated in terms of mean structural similarity values and normalized mean square error.
A present method was effective improvement of image quality can be expected by comparison with a Butterworth filter. Also, this method was considered clinical useful and a widely available simple method.

Free Research Field

放射線科学・医療画像情報解析学

Academic Significance and Societal Importance of the Research Achievements

術者が経験や知識に基づいて任意の処理条件設定を行う必要がないウェーブレット変換を利用した開発手法（Weighted Bayes Wavelet Shtinkage method）は、核医学画像における量子雑音を局所的、効率的かつ簡便に除去することが可能であり、一般臨床核医学検査における定量性、再現性、および標準化に大きく貢献する可能性が示唆された。また、医療被ばくの最適化や更なる医療被ばく線量の低減にも貢献することが示唆された。