Automated detection and tracking of abnormal regions on FDG-PET images

2014 Fiscal Year Final Research Report

• Project/Area Number: 24500543
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Medical systems
• Research Institution: Gifu University
• Principal Investigator: HARA Takeshi 岐阜大学, 医学(系)研究科(研究院), 准教授 (10283285)
• Co-Investigator(Kenkyū-buntansha): KATAFUCHI Tetsuro 岐阜医療科学大学, 保健科学部, 教授 (00393231) ; FUJITA Hiroshi 岐阜大学, 医学系研究科, 教授 (10124033)
• Co-Investigator(Renkei-kenkyūsha): ITO Satoshi 大雄会総合病院, 放射線科, 医長
• Project Period (FY): 2012-04-01 – 2015-03-31
• Keywords: 核医学 / FDG-PET / コンピュータ支援診断

Outline of Final Research Achievements

We have developed a prototype of computer-aided diagnosis system for FDG-PET images based on statistical image analysis. Fifty-five cases were employed to evaluate the initial performance. As a result of this work, we have achieved 95% true-positive fraction to detect abnormal spots and to track the activity changes between two time series examinations. The evaluation of activity changes of 76% examinations showed the same results as radiologists' comments.
In conclusions, the presentation of activity changes with SUV and Z-score will be a helpful index for the interpretation of FDG-PET images during the cancer treatment.

Free Research Field

医用画像解析