Abstract
Extraction of surface models of a hip joint from CT data is a pre-requisite step for computer assisted diagnosis and planning (CADP) of periacetabular osteotomy (PAO). Most of existing CADP systems are based on manual segmentation, which is time-consuming and hard to achieve reproducible results. In this paper, we present a Fully Automatic CT Segmentation (FACTS) approach to simultaneously extract both pelvic and femoral models. Our approach works by combining fast random forest (RF) regression based landmark detection, multi-atlas based segmentation, with articulated statistical shape model (aSSM) based fitting. The two fundamental contributions of our approach are: (1) an improved fast Gaussian transform (IFGT) is used within the RF regression framework for a fast and accurate landmark detection, which then allows for a fully automatic initialization of the multi-atlas based segmentation; and (2) aSSM based fitting is used to preserve hip joint structure and to avoid penetration between the pelvic and femoral models. Taking manual segmentation as the ground truth, we evaluated the present approach on 30 hip CT images (60 hips) with a 6-fold cross validation. When the present approach was compared to manual segmentation, a mean segmentation accuracy of 0.40, 0.36, and 0.36 mm was found for the pelvis, the left proximal femur, and the right proximal femur, respectively. When the models derived from both segmentations were used to compute the PAO diagnosis parameters, a difference of 2.0 ± 1.5°, 2.1 ± 1.6°, and 3.5 ± 2.3% were found for anteversion, inclination, and acetabular coverage, respectively. The achieved accuracy is regarded as clinically accurate enough for our target applications.
Similar content being viewed by others
References
Aljabar, P., R. Heckemann, A. Hammers, J. Hajnal, and D. Rueckert. Multi-atlas based segmentation of brain images: atlas selection and its effect on accuracy. NeuroImage 46(3):726–738, 2009.
Boykov, Y., O. Veksler, and R. Zabih. Fast approximate energy minimization via graph cuts. IEEE Trans. Pattern Anal. Mach. Intell. 23(11):1222–1239, 2001.
Breiman, L. Random forests. Mach. Learn. 45(1):5–32, 2001.
Cheng, Y., S. Zhou, Y. Wang, C. Guo, J. Bai, and S. Tamura. Automatic segmentation technique for acetabulum and femoral head in CT images. Pattern Recogn. 46(11):2969–2984, 2013.
Chu, C., M. Oda, T. Kitasaka, K. Misawa, M. Fujiwara, Y. Hayashi, Y. Nimura, D. Rueckert, and K. Mori. Multi-organ segmentation based on spatially-divided probabilistic atlas from 3D abdominal CT images. MICCAI 2013, Part II, pp. 165–172, 2013.
Criminisi, A., J. Shotton, D. Robertson, and E. Konukoglu. Regression forests for efficient anatomy detection and localization in CT studies. MCV 2010:106–117, 2010.
Ehrhardte, J., H. Handels, W. Plotz, and S. J. Poppl. Atlas-based recognition of anatomical structures and landmarks and the automatic computation of orthopedic parameters. Methods Inf. Med. 43(4):391–397, 2004.
Ganz, R., K. Klaue, T. Vinh, and J. Mast. A new periacetabular osteotomy for the treatment of hip dysplasia: technique and preliminary results. Clin. Orthop. 232:26–36, 1988.
Glocker, B., N. Komodakis, G. Tziritas, N. Navab, and N. Paragios. Dense image registration through MRFs and efficient linear programming. Med. Image Anal. 12(6):731–741, 2008.
Haas, B., T. Coradi, M. Scholz, P. Kunz, M. Huber, U. Oppitz, L. André, V. Lengkeek, D. Huyskens, A. van Esch, and R. Reddick. Automatic segmentation of thoracic and pelvic CT images for radiotherapy planning using implicit anatomic knowledge and organ-specific segmentation strategies. Phys. Med. Biol. 53(6):1751–1771, 2008.
Kainmueller, D., H. Lamecker, S. Zachow, and H.-C. Hege. An articulated statistical shape model for accurate hip joint segmentation. IEEE EMBC, pp. 6345–6351, 2009.
Kainmueller, D., H. Lamecker, S. Zachow, and H. C. Hege. Coupling deformable models for multi-object segmentation. ISBMS 2008:69–78, 2008.
Kang, Y. K. Engelke, and W. and A. Kalender. A new accurate and precise 3D segmentation method for skeletal structures in volumetric CT data. IEEE Trans. Med. Imaging 22(5):586–598, 2003.
Lamecker, H., M. Seebaß, H. C. Hege, and P. Deuflhard. A 3D statistical shape model of the pelvic bone for segmentation. SPIE 5370:1341–1351, 2004.
Lindner, C. S. Thiagarajah, J. M. Wilkinson, arcOGEN Consortium, G. Wallis, and T. F. Cootes. Fully automatic segmentation of the proximal femur using random forest regression voting. IEEE Trans. Med. Imaging 32(8):1462–1472, 2013.
Liu, L., T. Ecker, S. Schumann, K. Siebenrock, L.-P. Nolte, and G. Zheng. A novel planning and navigation system for peri-acetabular osteotomy (PAO). Int. J. Comput. Assist. Radiol. Surg. 9(Suppl 1):S168–S169, 2014.
McKinley, T. O. The Bernese Periacetabular Osteotomy: review of Reported Outcomes and the Early Experience at the University of Iowa. Iowa Orthop. J. 23:23–28, 2003.
Murray, D. W. The definition and measurement of acetabular orientation. J. Bone Joint Surg. [Br.] 75-B:228–232, 1993.
Pettersson, J., H. Knutsson, and M. Borga. Automatic Hip Bone Segmentation Using Non-Rigid Registration. ICPR 2006:946–949, 2006.
Seim, H., D. Kainmueller, M. Heller, H. Lamecker, S. Zachow, and H. C. Hege. Automatic segmentation of the pelvic bones from CT data based on a statistical shape model. VCBM 2008:93–100, 2008.
Vercauteren, T., X. Pennec, A. Perchant, and N. Ayache. Diffeomorphic demons: efficient non-parametric image registration. NeuroImage 45(1):61–72, 2009.
Viola, P., and M. Jones. Rapid object detection using a boosted cascade of simple features. CVPR 2001, vol. I, pp. 511–518, 2001.
Wolz, R., C. Chu, K. Misawa, M. Fujiwara, K. Mori, and D. Rueckert. Automated abdominal multi-organ segmentation with subject-specific atlas generation. IEEE Trans. Med. Imaging 32(9):1723–1730, 2013.
Xia, Y., J. Fripp, S. S Chandra, R. Schwarz, C. Engstrom and S. Crozier. Automated bone segmentation from large field of view 3D MR images of the hip joint. Phys. Med. Biol. 58:7375–7390.
Yang, C., R. Duraiswami, and L. Davis. Efficient Kernel Machines Using the Improved Fast Gauss Transform. NIPS 2005:1561–1568, 2005.
Yokota, F., T. Okada, M. Takao, S. Sugano, Y. Tada, N. Tomiyama, and Y. Sato. Automated CT segmentation of diseased hip using hierarchical and conditional statistical shape models. MICCAI 2013, Part II, pp. 190–197, 2013.
Yokota, F., T. Okada, M. Takao, S. Sugano, Y. Tada, and Y. Sato. Automated segmentation of the femur and pelvis from 3D CT data of diseased hip using hierarchical statistical shape model of joint structure. MICCAI 2009, Part II, pp. 811–818, 2009.
Zheng, G., X. Dong, K. Rajamani, X. Zhang, M. Styner, R. Thoranaghatte, and L.-P. Nolte. MA Gonzalez Ballester. Accurate and robust reconstruction of a surface model of the proximal femur from sparse-point data and a dense-point distribution model for surgical navigation. IEEE Trans. Biomed. Eng. 54:2109–2122, 2007.
Zoroofi, R. A., Y. Sato, T. Sasama, T. Nishii, N. Sugano, K. Yonenobu, H. Yoshikawa, T. Ochi, and S. Tamura. Automated segmentation of acetabulum and femoral head from 3-D CT images. IEEE Trans. Inf. Technol. Biomed. 7(4):329–343, 2003.
van Ginneken, B., T. Heimann, and M. Styner. 3D Segmentation in the Clinic: A Grand Challenge. In: T. Heimann, M. Styner, B. van Ginneken (Eds.): 3D Segmentation in the Clinic: A Grand Challenge, pp. 7-15, 2007.
Acknowledgments
The paper is partially supported by the Japanese-Swiss Science and Technology Cooperation Program and the Swiss National Science Foundation Project No. 205321_138009/1.
Conflict of interest
The authors have no conflict of interest related to this work.
Author information
Authors and Affiliations
Corresponding author
Additional information
Associate Editor Sean S. Kohles oversaw the review of this article.
Appendix
Appendix
How to convert Volumetric Overlap Error (OE) to Dice Overlap Coefficient (DOC)?
Let \(L_{1}\), \(L_{2}\) \(\subset {\mathbb{R}}^{3}\) denote the ground truth segmentation and the automatic segmentation, OE is calculated with the definition described in30:
And DOC is defined as
Since we have
After some mathematic manipulations, one can derive following relationship between OE and DOC:
For example, in Ref.,11 Kainmueller et al. reported an average OE of 9.7% for segmenting right hip bone (RHB) (Table IV of Ref. 11). The corresponding DOC is then 94.90%.
Rights and permissions
About this article
Cite this article
Chu, C., Chen, C., Liu, L. et al. FACTS: Fully Automatic CT Segmentation of a Hip Joint. Ann Biomed Eng 43, 1247–1259 (2015). https://doi.org/10.1007/s10439-014-1176-4
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10439-014-1176-4