Statistical Techniques for the Prediction of Vertebral Osteoporosis-Related Fractures

In this paper the problem of detection of osteoporosis related fractures is considered. Osteoporosis is a disease of the skeleton which is connected to the fragility of bones due to a loss of calcium. It is a disease which has various social and economical consequences and therefore early detection of osteoporotic fractures is a key issue for the health care system. The method introduced in this paper uses a set of non-invasive features to diagnose a vertebral fracture in women in postmenopausal phase. The level of recognition of vertebral fractures turns out to be quite satisfactory and suggests that the method can be used as an early screening technique.