Segmentation of the proximal femur in radial MR scans using a random forest classifier and deformable model registration
- 129 Downloads
Radial 2D MRI scans of the hip are routinely used for the diagnosis of the cam type of femoroacetabular impingement (FAI) and of avascular necrosis (AVN) of the femoral head, both considered causes of hip joint osteoarthritis in young and active patients. A method for automated and accurate segmentation of the proximal femur from radial MRI scans could be very useful in both clinical routine and biomechanical studies. However, to our knowledge, no such method has been published before.
The aims of this study are the development of a system for the segmentation of the proximal femur from radial MRI scans and the reconstruction of its 3D model that can be used for diagnosis and planning of hip-preserving surgery.
The proposed system relies on: (a) a random forest classifier and (b) the registration of a 3D template mesh of the femur to the radial slices based on a physically based deformable model. The input to the system are the radial slices and the manually specified positions of three landmarks. Our dataset consists of the radial MRI scans of 25 patients symptomatic of FAI or AVN and accompanying manual segmentation of the femur, treated as the ground truth.
The achieved segmentation of the proximal femur has an average Dice similarity coefficient (DSC) of 96.37 ± 1.55%, an average symmetric mean absolute distance (SMAD) of 0.94 ± 0.39 mm and an average Hausdorff distance of 2.37 ± 1.14 mm. In the femoral head subregion, the average SMAD is 0.64 ± 0.18 mm and the average Hausdorff distance is 1.41 ± 0.56 mm.
We validated a semiautomated method for the segmentation of the proximal femur from radial MR scans. A 3D model of the proximal femur is also reconstructed, which can be used for the planning of hip-preserving surgery.
KeywordsRadial imaging of the hip Proximal femur 3D reconstruction Segmentation Random forest Deformable model
This study was funded by the Swiss National Science Foundation (Grant number 205321_163224).
Compliance with ethical standards
Conflict of interest
All the authors declare that they have no conflict of interest.
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.
Informed consent was obtained from all individuals included in the study.
- 2.Sullivan JP, Griffith TB, Park CN, Ranawat AS (2017) Advances in 2D and 3D imaging for FAI surgical planning. In: Hip joint restoration. Springer, New York, pp 277–285Google Scholar
- 7.Lerch TD, Todorski IA, Steppacher SD, Schmaranzer F, Werlen SF, Siebenrock KA, Tannast M (2018) Prevalence of femoral and acetabular version abnormalities in patients with symptomatic hip disease: a controlled study of 538 hips. Am J Sports Med 46(1):122–134. https://doi.org/10.1177/0363546517726983 CrossRefPubMedGoogle Scholar
- 17.Schmaranzer F, Todorski IAS, Lerch TD, Schwab J, Cullmann-Bastian J, Tannast M (2017) Intra-articular lesions: imaging and surgical correlation. In: Seminars in musculoskeletal radiology, vol 21, No. 05. Thieme Medical Publishers, pp 487–506Google Scholar
- 20.Lerch T, Degonda C, Zheng G, Todorski I, Schmaranzer F, Ecker T, Siebenrock K, Tannast M (2017) MR-based 3D PAO planning and simulation of hip impingement is as accurate as CT-based 3D models. German Congress of Orthopedic and Trauma Surgery (DKOU 2017)Google Scholar
- 26.Zeng G, Yang X, Li J, Yu L, Heng PA, Zheng G (2017) 3D U-net with multi-level deep supervision: fully automatic segmentation of proximal femur in 3D MR images. In: International workshop on machine learning in medical imaging. Springer, Cham, pp 274–282Google Scholar
- 29.Tu Z (2008) Auto-context and its application to high-level vision tasks. In: IEEE conference on computer vision and pattern recognition, CVPR 2008. IEEE, pp 1–8Google Scholar
- 30.Gao Y, Wang L, Shao Y, Shen D (2014) Learning distance transform for boundary detection and deformable segmentation in ct prostate images. In: International workshop on machine learning in medical imaging. Springer, Cham, pp 93–100Google Scholar
- 32.Glocker B, Zikic D, Konukoglu E, Haynor DR, Criminisi A (2013) Vertebrae localization in pathological spine CT via dense classification from sparse annotations. In: International conference on medical image computing and computer-assisted intervention. Springer, Berlin, pp 262–270Google Scholar
- 34.Schmid J, Magnenat-Thalmann N (2008) MRI bone segmentation using deformable models and shape priors. In: International conference on medical image computing and computer-assisted intervention. Springer, Berlin, pp 119–126Google Scholar
- 35.Volino P, Magnenat-Thalmann N (2000) Implementing fast cloth simulation with collision response. In: Proceedings of the computer graphics international. IEEE, pp 257–266Google Scholar
- 36.Cootes TF, Hill A, Taylor CJ, Haslam J (1993) The use of active shape models for locating structures in medical images. In: Biennial international conference on information processing in medical imaging. Springer, Berlin, pp 33–47Google Scholar
- 38.Kumar S (2003) Discriminative random fields: a discriminative framework for contextual interaction in classification. In: Proceedings of the 9th IEEE international conference on computer vision, 2003. IEEE, pp 1150–1157Google Scholar
- 42.Zikic D, Glocker B, Konukoglu E, Criminisi A, Demiralp C, Shotton J, Thomas OM, Das T, Jena R, Price SJ (2012) Decision forests for tissue-specific segmentation of high-grade gliomas in multi-channel MR. In: International conference on medical image computing and computer-assisted intervention. Springer, Berlin, pp 369–376Google Scholar
- 44.Montillo A, Shotton J, Winn J, Iglesias JE, Metaxas D, Criminisi A (2011) Entangled decision forests and their application for semantic segmentation of CT images. In: Biennial international conference on information processing in medical imaging. Springer, Berlin, pp 184–196Google Scholar
- 47.Louppe G (2014) Understanding random forests: from theory to practice. arXiv Preprint arXiv:1407.7502
- 48.Criminisi A, Shotton J (eds) (2013) Decision forests for computer vision and medical image analysis. Springer, BerlinGoogle Scholar
- 49.Damopoulos D, Glocker B, Zheng G (2017) Automatic localization of the lumbar vertebral landmarks in CT images with context features. In: International workshop and challenge on computational methods and clinical applications in musculoskeletal imaging. Springer, Cham, pp 59–71Google Scholar
- 51.Kosov S (2013) Direct graphical models C++ library. http://research.project-10.de/dgm/