Women with previous fragility fractures can be classified based on bone microarchitecture and finite element analysis measured with HR-pQCT
High-resolution peripheral quantitative computed tomography (HR-pQCT) measurements of distal radius and tibia bone microarchitecture and finite element (FE) estimates of bone strength performed well at classifying postmenopausal women with and without previous fracture. The HR-pQCT measurements outperformed dual energy x-ray absorptiometry (DXA) at classifying forearm fractures and fractures at other skeletal sites.
Areal bone mineral density (aBMD) is the primary measurement used to assess osteoporosis and fracture risk; however, it does not take into account bone microarchitecture, which also contributes to bone strength. Thus, our objective was to determine if bone microarchitecture measured with HR-pQCT and FE estimates of bone strength could classify women with and without low-trauma fractures.
We used HR-pQCT to assess bone microarchitecture at the distal radius and tibia in 44 postmenopausal women with a history of low-trauma fracture and 88 age-matched controls from the Calgary cohort of the Canadian Multicentre Osteoporosis Study (CaMos) study. We estimated bone strength using FE analysis and simulated distal radius aBMD from the HR-pQCT scans. Femoral neck (FN) and lumbar spine (LS) aBMD were measured with DXA. We used support vector machines (SVM) and a tenfold cross-validation to classify the fracture cases and controls and to determine accuracy.
The combination of HR-pQCT measures of microarchitecture and FE estimates of bone strength had the highest area under the receiver operating characteristic (ROC) curve of 0.82 when classifying forearm fractures compared to an area under the curve (AUC) of 0.71 from DXA-derived aBMD of the forearm and 0.63 from FN and spine DXA. For all fracture types, FE estimates of bone strength at the forearm alone resulted in an AUC of 0.69.
Models based on HR-pQCT measurements of bone microarchitecture and estimates of bone strength performed better than DXA-derived aBMD at classifying women with and without prior fracture. In future, these models may improve prediction of individuals at risk of low-trauma fracture.
KEYWORDSBone microarchitecture Finite element analysis Fracture HR-pQCT Support vector machines
- 8.Melton LJ, Riggs BL, van Lenthe GH, Achenbach SJ, Müller R, Bouxsein ML, Amin S, Atkinson EJ, Khosla S (2007) Contribution of in vivo structural measurements and load/strength ratios to the determination of forearm fracture risk in postmenopausal women. J Bone Miner Res 22:1442–1448PubMedCrossRefGoogle Scholar
- 12.Atkinson EJ, Therneau TM, Melton LJ, Camp JJ, Achenbach SJ, Amin S and Khosla S (2012) Assessing fracture risk using gradient boosting machine (GBM) models. J Bone Miner ResGoogle Scholar
- 27.Vapnik V Estimation of dependences based on empirical data: Springer series in statistics. Springer, New YorkGoogle Scholar
- 28.Hall M, Frank E, Holmes G, Pfahringer B, Reutemann P, Witten IH (2009) The WEKA data mining software: an update. SIGKDD Explorations 11Google Scholar
- 29.Cortes C, Vapnik V (1995) Support-vector networks. Mach Learn 20:273–297Google Scholar
- 30.Valyon J, Horváth G (2003) A weighted generalized LS-SVM. Period Polytech Electr Eng 47:229–251Google Scholar
- 32.Harvey N, Dennison E, Cooper C, ASBMR (2009) Chapter 38. Epidemiology of osteoporotic fractures. In: Rosen JC (ed) Primer on the metabolic bone diseases and disorders of mineral metabolism, 7th edn. American Society for Bone and Mineral Research, Washington, pp 197–203Google Scholar