An automatic segmentation system of acetabulum in sequential CT images for the personalized artificial femoral head design. Issue 127 (April 2016)

• Record Type: Journal Article
• Authors: Liu, Bin; Zhang, Hui; Hua, Shungang; Jiang, Qianfeng; Huang, Rui; Liu, Wenpeng; Zhang, Shuo; Zhang, Bingbing; Yue, Zongge
• Is Part Of: Computer methods and programs in biomedicine. Issue 127(2016)
• Page Start: 318
• Page End: 335
• Publication Date: 2016-04
• Subjects: Acetabulum segmentation -- Personalized modeling -- Artificial femoral head -- Sequential CT images -- Snake model -- Avascular necrosis of femoral head; Medicine -- Computer programs -- Periodicals; Biology -- Computer programs -- Periodicals; Computers -- Periodicals; Medicine -- Periodicals; Médecine -- Logiciels -- Périodiques; Biologie -- Logiciels -- Périodiques; Biology -- Computer programs; Medicine -- Computer programs; Periodicals; Electronic journals; 610.28
• DOI: 10.1016/j.cmpb.2015.12.012
• Languages: English
• ISSNs: 0169-2607

Abstract

Highlights: We develop an automatic method for segmenting the acetabulum from CT images. We can obtain the accurate sequential contours of the acetabulum. 3D models of the acetabulum obtained by our method are accurate and satisfactory. We can reconstruct the personalized femoral head 3D prosthesis models. Abstract: This paper describes an automatic and accurate segmentation method to extract the acetabulum tissue from sequential CT images. The hip joint consists of acetabulum and femoral head. In the personalized femoral head prosthesis designing by reverse engineering technology, obtaining the accurate acetabulum shape is the most important task. However, due to the necrotic femoral head's complex shape and the extremely narrow inter-bone region, obtaining the accurate acetabulum shape remains a challenging work. In this paper, we overcame these difficulties and developed an automatic segmentation method. First, we obtain the rough contour of the femoral head by utilizing the constraints of the great trochanter and the shape of femoral head in the initial slice. Second, we refine the rough contour by an orthogonal line edge detection approach and obtain a refined contour which will be used as the initial contour of the snake algorithm. Then, the snake algorithm is performed slice by slice upwards and downwards to generate the adjacent contours. During this process, the contour of the femoral head in a segmented slice is used as the initial contour of the next unsegmented slice. Finally, we can obtain the accurate sequential contours of the acetabulum by removing the femoral head and the femoral regions. And the 3D models of the acetabulum can be obtained correspondingly. The experimental result shows that the 3D models obtained by the proposed method are accurate and satisfactory. On this condition, we can reconstruct the personalized femoral head 3D models and design the personalized femoral head prosthesis.