The Third Computational Anatomy Seminar



  • Presenter
    • Kenji Suzuki, Ph.D (Assistant Professor of Radiology, Medical Physics, and Comprehensive Cancer Center, Department of Radiology, Division of the Biological Sciences, The University of Chicago )
  • Title
    • Machine-Learning Approach to Medical Pattern Recognition
  • Abstract
    • Machine leaning plays an essential role in medical pattern recognition, because objects in medical images such as lesions and anatomic structures cannot be represented by simple equations accurately; thus, tasks in medical pattern recognition require “learning from examples” essentially. We have been studying on machine-learning methods called massive-training artificial neural networks (MTANNs) for various tasks in medical pattern recognition. An MTANN is a supervised pixel/voxel-based machine-learning technique for medical image processing and pattern recognition. The MTANN directly learns the relationship between “teaching” input and its desired images to enhance specific patterns and suppresses other patterns in medical images. An MTANN is a versatile tool; and thus, it is applicable to various tasks in medical pattern recognition, such as enhancing edges traced by physicians, lesion enhancement on CT images, lung nodule detection on thoracic CT images and chest radiographs, distinction between benign and malignant nodules on CT images, separation of ribs from soft tissue in chest radiographs, and polyp detection in CT colonography


  • Yoshinobu Sato, PhD (Associate Professor, Department of Radiology, Osaka University)
    • Phone +81 6 6879 3562
    • E-mail yoshi @

Cooperating Program

トップ   編集 差分 バックアップ 添付 複製 名前変更 リロード   新規 一覧 単語検索 最終更新   ヘルプ   最終更新のRSS
Last-modified: 2013-09-03 (Tue) 11:24:15 (1507d)