Automated Detection Method of Breast Tumors on Cross-Sectional Ultrasonograms
スポンサーリンク
概要
- 論文の詳細を見る
We have developed a computer-aided detection scheme of breast tumors on cross-sectional ultrasonograms. The scheme proposed in this study consists of the following three steps: (1)region segmentation by using a newly devised erosion and dilation processing with distance transformation, (2) feature analysis by using shape of tumor, and (3) three-dimensional (3D) display for detection output. Our method was found to be effective for the detection of tumors examined by 9 cases including 459 images. The sensitivity of our scheme was 78% with 4.2 false-positive (FP) findings per case. The reducation of the FP candidates is required for future works.
- Medical Imaging and Information Sciencesの論文
Medical Imaging and Information Sciences | 論文
- 系列依存性による2次元画像の解析およびその周辺
- 小型X線マイクロビームスキャナによるディジタル画像
- EMIスキヤン像のデータ処理-アナログメモリを用いた多重ウィンドウ表示と多重スライス表示-
- RadiographにおけるGraininessとSharpnessの検討
- X線用イメジインテンシファイアの雑音について