[Zurück]

P. Varga, S. Baumbach, D. H. Pahr, P.K. Zysset:
"Finite Element Models of Human Radius Slices can Accurately Predict Colles' Fracture Load";
Poster: 36th European Symposium on Calcified Tissues, Wien; 23.05.2009 - 27.05.2009.

Distal radius fractures occur prior to other osteoporotic fractures and may be precursors to identify patients at risk early in time. High resolution peripheral quantitative computer tomography (HR-pQCT) provides in vivo access to the trabecular architecture of the peripheral skeleton. Following successful in vitro validation studies, HR-pQCT based anatomy specific finite element (FE) models may succeed to densitometry and morphological indices in predicting fracture risk.
In a previous study, an experimental model of Colles' fracture was developed. 21 embalmed radius bones were cut to 1/3 length, scanned using HR-pQCT and tested mechanically in compression until failure with well defined boundary conditions. The goals of the present study were to evaluate 1) if a distal slice of the human radius provides an adequate model to estimate full bone strength and 2) if smooth surface based patient specific continuum FE models can indeed predict strength better than morphological indices.
Two distinct 9 mm thick image sections of each radius were selected for analysis: one corresponding to the region of the clinical protocol and a second one, ultradistal, located adjacent to the endplate. Standard morphological indices were evaluated using the software of the HR-pQCT system. Continuum FE models were created from the HR-pQCT images separating spongious and cortical regions and assigning material properties based on local volume fraction (BV/TV) and fabric. Nonlinear analyses of axial compression of the slices were performed using a constitutive law including elasticity, plasticity and damage of which all constants were taken from previous experiments of biopsies.
Exceptional prediction of the experimental fracture load was achieved using the FE method (adj. R2 = 0.93, p < 0.0001). Correlation between the experimental strength and all the investigated measures were higher for the most distal region compared to the standard section. As expected, the FE method provided a superior prediction of fracture load compared to morphological parameters such as trabecular BV/TV (R2 = 0.82) or cortical thickness (R2 = 0.26) as it integrates the most relevant of them in a mechanical instead of a statistical way. These results underline the outstanding capabilities of the HR-pQCT based FE method and support the use of a 9 mm wrist section in the clinical evaluation of fracture risk. They also suggest the need for shifting the region of analysis distally, adjacent to the endplate.

Projektleitung Philippe K. Zysset:
Voraussage des Bruchrisikos beim distalen Radius