Image analysis in ultrasonography using wavelet transform and its application to the diagnosis of Sjoegren's syndrome

2016 Fiscal Year Final Research Report

• Project/Area Number: 26462843
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Pathobiological dentistry/Dental radiology
• Research Institution: Kyushu University
• Principal Investigator: Ohki Masafumi 九州大学, 医学研究院, 教授 (10160441)
• Co-Investigator(Renkei-kenkyūsha): NAKAMURA Takashi 長崎大学, 大学院医歯薬学総合研究科, 教授 (30172406)
• Research Collaborator: MATSUI Takehiro
• Project Period (FY): 2014-04-01 – 2017-03-31
• Keywords: ウェーブレット変換 / シェーグレン症候群 / 超音波画像診断

Outline of Final Research Achievements

In this study, we aimed to develop an image analysis for improving diagnostic accuracy in ultrasonography, which is highly important as a noninvasive diagnosis of Sjoegren's syndrome. We developed an image analysis using dual tree complex wavelet transform (DT - CWT) and analyzed the parotid gland ultrasonographic images of 174 patients, including 77 cases who were considered positive for Sjoegren 's syndrome by sialography. As a result of the diagnostic accuracy combined with DT-CWT and machine learning, the sensitivity in the low severity group was 90 ± 3.7% and the accuracy was 88 ± 4.2%. The fact that the sensitivity was 79% and the accuracy was 80% by a dentist indicated the usefulness of this image analysis in clinical diagnosis.

Free Research Field

医用画像処理