Osteoporosis: sensitivity and specificity of dental radiographs

Mar 22 2007

Researchers in the Academic Center for Dentistry Amsterdam have created a unique way of identifying patients at risk of osteoporosis by using ordinary dental x-rays. Professor Paul F. van der Stelt and his team developed the largely automated approach to detecting the disease during a three-year, EU-funded collaboration with the Universities of Manchester, Athens, Leuven, and Malmo.

They will present their findings today during the 85th General Session of the International Association for Dental Research.

Osteoporosis affects almost 15% of Western women in their fifties, 22% in their sixties, and 38.5% in their seventies. As many as 70% of women over 80 are at risk, and the condition carries a high risk of bone fractures, with over a third of adult women falling victim at least once in their lifetime. Wide-scale screening for the disease is not currently viable, largely due to the cost and scarcity of specialist equipment and staff.

The team has therefore developed an innovative software-based approach to detecting osteoporosis using routine dental x-rays, by automatically analyzing specific characteristics of the radiographic trabecular bone pattern. These features include, among others, the thickness, the amount of fragmentation, and the main orientation of the structure of the trabecular bone.

In four clinical centers, 671 women with an average age of 55 years were recruited. To obtain the "gold standard", the team measured bone thickness at the femur, hip, and spine, using the technique that is common for this kind of expensive examination (Bone Mass Density, BMD). In addition, one panoramic and two intra-oral radiographs were made.

X-rays are used widely in dental practice for several reasons. Using the image information from these radiographs to detect patients at risk for osteoporosis involves no extra radiation and almost no extra cost, while undetected osteoporotic patients can incur bone fractures and suffer from other problems, reducing the quality of life.

By analyzing a small area depicting the trabecular bone on the ordinary dental radiographs, dentists can predict the osteoporotic condition of the patients to the same extent as the BMD measurements.

The image analysis can be automated and does not require extra time on the part of the dentist. The newly developed technique means that patients who are identified as being at risk can be referred for more thorough appropriate examinations.

These findings are published online by the Elsevier journal Bone and will appear in print soon.

This work was supported by a research and technological development project grant from the European Commission Fifth Framework Program, 'Quality of Life and Management of Living Resources', QLK6-2002-02243, 'Osteodent'.

http://www.iadr.org