脳MR画像におけるラクナ梗塞の検出法の改良-AdaBoostテンプレートマッチングを用いた偽陽性削除-  [in Japanese] Improvement of Automatic Detection Method of Lacunar Infarcts on MR Images: Reduction of False Positives By Using AdaBoost Template Matching  [in Japanese]

Access this Article

Search this Article

Author(s)

    • 内山 良一 UCHIYAMA Yoshikazu
    • 熊本大学大学院生命科学研究部先端生命医療科学部門 Department of Medical Physics, Faculty of Life Science, Kumamoto University
    • 村松 千左子 MURAMATSU Chisako
    • 岐阜大学大学院医学系研究科知能イメージ情報分野 Department of Intelligent Image Information, Graduate School of Medicine, Gifu University
    • 原 武史 HARA Takeshi
    • 岐阜大学大学院医学系研究科知能イメージ情報分野 Department of Intelligent Image Information, Graduate School of Medicine, Gifu University
    • 白石 順二 SHIRAISHI Junji
    • 熊本大学大学院生命科学研究部先端生命医療科学部門 Department of Medical Physics, Faculty of Life Science, Kumamoto University
    • 藤田 広志 FUJITA Hiroshi
    • 岐阜大学大学院医学系研究科知能イメージ情報分野 Department of Intelligent Image Information, Graduate School of Medicine, Gifu University

Abstract

The detection of lacunar infarcts is important because their presence indicates an increased risk of severe cerebral infarction. However, their accurate identification is often hard because of the difficulty in distinguishing between lacunar infarcts and enlarged Virchow-Robin spaces. Therefore, we developed computer-aided diagnosis scheme for the detection of lacunar infarcts. The performance of our previous method indicated that the sensitivity of 96.8% with 0.76 false positive(FP)per slice. However, further reduction of FPs was remained as an issue to be solved for the clinical application. In this paper, we proposed AdaBoost template matching. This classifier can distinguish between lacunar infarcts and FPs by selecting suitable templates in the template matching. By using this technique, 55.5% FPs were eliminated while keeping the same sensitivity. Thus the proposed method was found to be useful for the sophistication of the automatic detection of lacunar infarcts in MR images.

Journal

  • Medical Imaging and Information Sciences

    Medical Imaging and Information Sciences 31(2), 41-46, 2014

    MEDICAL IMAGING AND INFORMATION SCIENCES

Codes

  • NII Article ID (NAID)
    130004876122
  • NII NACSIS-CAT ID (NCID)
    AN10156808
  • Text Lang
    JPN
  • Article Type
    journal article
  • ISSN
    0910-1543
  • Data Source
    IR  J-STAGE 
Page Top