Vertebral body insufficiency fractures: detection of vertebrae at risk on standard CT images using texture analysis and machine learning
To evaluate the diagnostic performance of bone texture analysis (TA) combined with machine learning (ML) algorithms in standard CT scans to identify patients with vertebrae at risk for insufficiency fractures.
Materials and methods
Standard CT scans of 58 patients with insufficiency fractures of the spine, performed between 2006 and 2013, were analyzed retrospectively. Every included patient had at least two CT scans. Intact vertebrae in a first scan that either fractured (“unstable”) or remained intact (“stable”) in the consecutive scan were manually segmented on mid-sagittal reformations. TA features for all vertebrae were extracted using open-source software (MaZda). In a paired control study, all vertebrae of the study cohort “cases” and matched controls were classified using ROC analysis of Hounsfield unit (HU) measurements and supervised ML techniques. In a within-subject vertebra comparison, vertebrae of the cases were classified into “unstable” and “stable” using identical techniques.
One hundred twenty vertebrae were included. Classification of cases/controls using ROC analysis of HU measurements showed an AUC of 0.83 (95% confidence interval [CI], 0.77–0.88), and ML-based classification showed an AUC of 0.97 (CI, 0.97–0.98). Classification of unstable/stable vertebrae using ROC analysis showed an AUC of 0.52 (CI, 0.42–0.63), and ML-based classification showed an AUC of 0.64 (CI, 0.61–0.67).
TA combined with ML allows to identifying patients who will suffer from vertebral insufficiency fractures in standard CT scans with high accuracy. However, identification of single vertebra at risk remains challenging.
• Bone texture analysis combined with machine learning allows to identify patients at risk for vertebral body insufficiency fractures on standard CT scans with high accuracy.
• Compared to mere Hounsfield unit measurements on CT scans, application of bone texture analysis combined with machine learning improve fracture risk prediction.
• This analysis has the potential to identify vertebrae at risk for insufficiency fracture and may thus increase diagnostic value of standard CT scans.
KeywordsSpine Machine learning Osteoporosis Tomography, X-ray computed
Artificial neural networks
Bone mineral density
Concordance correlation coefficient
Dual-energy X-ray absorptiometry
Finite element analysis
Gray-level co-occurrence matrix
Gray-level run-length matrix
High-resolution peripheral quantitative computed tomography
Region of interest
Support vector machine
The authors state that this work has not received any funding.
Compliance with ethical standards
The scientific guarantor of this publication is Roman Guggenberger.
Conflict of interest
The authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article.
Statistics and biometry
One of the authors has significant statistical expertise.
Written informed consent was waived by the Institutional Review Board.
Institutional Review Board approval was obtained.
Study subjects or cohorts overlap
The control cohort used in the present study (58 patients) is part of a study population of a previously published study establishing normative values for CT-based texture analysis of vertebral bodies. Mannil M, Eberhard M, Becker AS, et al (2017) Normative values for CT-based texture analysis of vertebral bodies in dual X-ray absorptiometry-confirmed, normally mineralized subjects. Skeletal Radiology 46:1541–1551. https://doi.org/10.1007/s00256-017-2728-0).
• case-control study
• performed at one institution
- 8.Burns JE, Yao J, Summers RM (2017) Vertebral body compression fractures and bone density: automated detection and classification on CT images. Radiology 284:788–797. https://doi.org/10.1148/radiol.2017162100
- 9.Schreiber JJ, Anderson PA, Rosas HG, Buchholz AL, Au AG (2011) Hounsfield units for assessing bone mineral density and strength: a tool for osteoporosis management. J Bone Joint Surg 93:1057–1063. https://doi.org/10.2106/JBJS.J.00160
- 12.Imai K, Ohnishi I, Bessho M, Nakamura K (2006) Nonlinear finite element model predicts vertebral bone strength and fracture site. Spine (Phila Pa 1976) 31:1789–1794Google Scholar
- 14.Lubner MG, Smith AD, Sandrasegaran K, Sahani DV, Pickhardt PJ (2017) CT texture analysis: definitions, applications, biologic correlates, and challenges. Radiographics 37:1483–1503. https://doi.org/10.1148/rg.2017170056
- 15.Rachidi M, Marchadier A, Gadois C, Lespessailles E, Chappard C, Benhamou CL (2008) Laws’ masks descriptors applied to bone texture analysis: an innovative and discriminant tool in osteoporosis. Skeletal Radiol 37:541–548. https://doi.org/10.1007/s00256-008-0463-2
- 17.Zou Z, Yang J, Megalooikonomou V, Jennane R, Cheng E, Ling H (2016) Trabecular bone texture classification using wavelet leaders. Proc. SPIE 9788, Medical Imaging 2016: Biomedical Applications in Molecular, Structural, and Functional Imaging, 97880E. https://doi.org/10.1117/12.2216452
- 19.Tabari A, Torriani M, Miller KK, Klibanski A, Kalra MK, Bredella MA (2016) Anorexia nervosa: analysis of trabecular texture with CT. Radiology 283:178–185Google Scholar
- 26.Kuhn M (2008) Building predictive models in R using the caret package. J Stat Softw 28. https://doi.org/10.18637/jss.v028.i05
- 27.Gevrey M, Dimopoulos I, Lek S (2003) Review and comparison of methods to study the contribution of variables in artificial neural network models. Ecol Modell 160:249–264Google Scholar
- 29.Lin LI-K (1989) A concordance correlation coefficient to evaluate reproducibility. Biometrics 45:255–268. https://doi.org/10.2307/2532051
- 31.Valentinitsch A, Patsch J, Mueller D et al (2010) Texture analysis in quantitative osteoporosis assessment. In: Biomedical imaging: from nano to macro, 2010 IEEE International Symposium on. IEEE, pp 1361–1364Google Scholar
- 32.Boivin GY, Chavassieux PM, Santora AC, Yates J, Meunier PJ (2000) Alendronate increases bone strength by increasing the mean degree of mineralization of bone tissue in osteoporotic women. Bone 27:687–694Google Scholar
- 33.Guggenbuhl P, Bodic F, Hamel L, Baslé MF, Chappard D (2006) Texture analysis of X-ray radiographs of iliac bone is correlated with bone micro-CT. Osteoporos Int 17:447–454. https://doi.org/10.1007/s00198-005-0007-8
- 34.Chappard D, Guggenbuhl P, Legrand E, Baslé MF, Audran M (2005) Texture analysis of X-ray radiographs is correlated with bone histomorphometry. J Bone Miner Metab 23:24–29. https://doi.org/10.1007/s00774-004-0536-9
- 35.Kimmel DB, Recker RR, Gallagher JC, Vaswani AS, Aloia JF (1990) A comparison of iliac bone histomorphometric data in post-menopausal osteoporotic and normal subjects. Bone Miner 11:217–235Google Scholar
- 36.Reddy TK, Kumaravel N (2010) Wavelet based texture analysis and classification of bone lesions from dental CT. Int J Med Eng Inf 2:319–327Google Scholar
- 38.Paul R, Alahamri S, Malla S, Quadri GJ (2017) Make your bone great again: a study on osteoporosis classification. Available via http://arxiv.org/abs/1707.05385. Accessed 02 Jan 2018