講演抄録/キーワード |
講演名 |
2008-09-17 14:35
Semi-automated method for detecting hip joint contours from 2D X-ray images using statistical shape model ○Lassad Ben younes(Univ. Paris-12-ESIEE/Osaka Univ.)・Toshiyuki Okada(Osaka Univ. Hospital)・Takashi Nishii・Nobuhiko Sugano・Yoshinobu Sato(Osaka Univ.) MI2008-43 |
抄録 |
(和) |
To follow the progress of osteoarthritis (OA), degenerative joint disease, 2D X-ray images are taken. This paper describes a method which delimits, semi-automatically, the femoral head and acetabulum, inferior margin of pelvis. The algorithm requires that few points are, manually, selected: 3 on the acetabulum edge and 5 on the femoral head edge. The processing scheme consisted of the following step: 1) A statistical shape model(SSM) is created using 2D X-ray datasets of 33 patients; 2) Few edge points of the femoral head and acetabulum are specified by user; 3) The statistical shape model is fitting to selected points; 4) An iterative algorithm is applied which consist on; a) detection of boundary candidates, based on pixel intensity, along perpendicular direction of current SSM contours; b) fitting SSM to the detected boundary candidates. The algorithm was testing with 5 X-ray images. Two differents users select initial points. One of them is the Orthopaedic surgeon who segmented, manually, X-ray images for SSM. |
(英) |
To follow the progress of osteoarthritis (OA), degenerative joint disease, 2D X-ray images are taken. This paper describes a method which delimits, semi-automatically, the femoral head and acetabulum, inferior margin of pelvis. The algorithm requires that few points are, manually, selected: 3 on the acetabulum edge and 5 on the femoral head edge. The processing scheme consisted of the following step: 1) A statistical shape model(SSM) is created using 2D X-ray datasets of 33 patients; 2) Few edge points of the femoral head and acetabulum are specified by user; 3) The statistical shape model is fitting to selected points; 4) An iterative algorithm is applied which consist on; a) detection of boundary candidates, based on pixel intensity, along perpendicular direction of current SSM contours; b) fitting SSM to the detected boundary candidates. The algorithm was testing with 5 X-ray images. Two differents users select initial points. One of them is the Orthopaedic surgeon who segmented, manually, X-ray images for SSM. |
キーワード |
(和) |
Hip joint detection / Segmentation / Statistical shape model / Active shape model / X-ray images / / / |
(英) |
Hip joint detection / Segmentation / Statistical shape model / Active shape model / X-ray images / / / |
文献情報 |
信学技報, vol. 108, no. 209, MI2008-43, pp. 41-46, 2008年9月. |
資料番号 |
MI2008-43 |
発行日 |
2008-09-10 (MI) |
ISSN |
Print edition: ISSN 0913-5685 Online edition: ISSN 2432-6380 |
著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します.(許諾番号:10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
PDFダウンロード |
MI2008-43 |
|