Automated Classification of Cerebral Arteries in MRA Images

DOI Web Site 11 Citations 20 References
  • YAMAUCHI Masashi
    Department of Intelligent Image Information, Graduate School of Medicine, Gifu University
  • UCHIYAMA Yoshikazu
    Department of Intelligent Image Information, Graduate School of Medicine, Gifu University
  • OGURA Jun
    Department of Intelligent Image Information, Graduate School of Medicine, Gifu University
  • YOKOYAMA Ryujiro
    Department of Intelligent Image Information, Graduate School of Medicine, Gifu University
  • HARA Takeshi
    Department of Intelligent Image Information, Graduate School of Medicine, Gifu University
  • ANDO Hiromichi
    Department of Neurosurgery, Gifu Municipal Hospital
  • YAMAKAWA Hiroyasu
    Department of Neurosurgery, Prefecture Gero Hot Springs Hospital
  • IWAMA Toru
    Department of Neurosurgery, Graduate School of Medicine, Gifu University
  • HOSHI Hiroaki
    Department of Radiology, Graduate School of Medicine, Gifu University
  • FUJITA Hiroshi
    Department of Intelligent Image Information, Graduate School of Medicine, Gifu University

Bibliographic Information

Other Title
  • MRA画像における脳血管名の自動対応付け手法の開発
  • MRA ガゾウ ニ オケル ノウケッカンメイ ノ ジドウ タイオウズケ シュホウ ノ カイハツ

Search this article

Abstract

The detection of unruptured aneurysms is a major task in magnetic resonance angiography (MRA). However, it is difficult for radiologists and/or neurosurgeons to detect small aneurysms on maximum intensity projection (MIP) images because adjacent vessels may overlap with the aneurysms. Therefore, we proposed a method for making a new MIP image, the SelMIP image, containing interested vessels only by manually selecting a cerebral artery from a list of cerebral arteries recognized automatically. For the automated classification of cerebral arteries, two three-dimesional images, a target image and a reference image, were compared. Image registration was performed using global matching and rigidity transformation. The segmented vessel regions were classified into eight cerebral arteries by calculating the Euclidian distance between a voxel in the target image and each of the voxels in the eight labeled vessel regions in the reference image. In applying the automated cerebral arteries recognition algorithm to 110 MRA studies, the results of subjective evaluation were that 76.4% (84/110) were rated as good, 13.6% (15/110) as adequate, and 10.0% (11/110) as poor. The results rated good or adequate are considered acceptable and would be adequate for clinical use. Overall, 90.0% (99/110) of MRA studies attained a clinically acceptable result. Our new viewing technique will be useful in assisting radiologists to detectaneurysms and reducing the interpretation time.

Journal

Citations (11)*help

See more

References(20)*help

See more

Keywords

Details 詳細情報について

Report a problem

Back to top