Osteoporotic hip fractures represent a high risk for people over 60 years of age, with a
reduction in life expectancy of 20% and a reduction in functional capacity of 50%.
Currently, dual X-ray absorptiometry (DXA) is the gold standard for the diagnosis of
osteoporosis, although it has a low prediction capacity of bone fracture. Likewise, there
exist other techniques such as quantitative computed tomography measurements with a
better prediction of bone fracture, although limited in practice ...
Osteoporotic hip fractures represent a high risk for people over 60 years of age, with a
reduction in life expectancy of 20% and a reduction in functional capacity of 50%.
Currently, dual X-ray absorptiometry (DXA) is the gold standard for the diagnosis of
osteoporosis, although it has a low prediction capacity of bone fracture. Likewise, there
exist other techniques such as quantitative computed tomography measurements with a
better prediction of bone fracture, although limited in practice due to their high radiation
doses. On the other hand, finite element models (FEM) represent a tool that provides
complementary information, such as strength and stress, which can be better predictors
of hip fractures. Thus, the combination of advanced DXA images with FEM allowed a
better discrimination of fracture and non-fracture cases with high accuracy. However, the
application of these models in clinical practice is complicated due to the computational
time and data processing. Therefore, it would be of interest to find an alternative way to
calculate such parameters in a shorter time without the need to perform such FEM
simulations. With this in mind, the aim of this project is to explore statistical methods to
correlate DXA measurements and mechanical stress and strain parameters obtained by
FEM simulations. To carry out the study, it initially started from a database of 111 patients
including DXA measurements and mechanical parameters derived from previous FEM
simulations, which were subsequently expanded to balance the number of patients
between males and females. Afterwards, correlation analysis was performed to extract
the minimum significant DXA parameters for the prediction of mechanical parameters. It
was done using a SSPS statistical program. Likewise, a comparison between men and
women was also performed, as well as other types of regressions and a comparison
considering the type of hip fracture (distinguishing between neck and trochanter). The
study showed that the regression between the DXA measurements and the biomechanical
parameters of stress and strain were different between osteoporotic patients without
fracture and fractured patients because the structural changes that occur in the bone are
not only due to the amount of bone mineral density but also because they have different
mechanical behavior from one case to another.
+
