A Computer Aided System to Measure Mandibular Cortical Width on Dental Radiograph in Prediction of Osteoporosis

Abstract

Osteoporosis is a musculoskeletal disorder affecting the population in millions all over the globe. It is an epidemic and mainly under diagnosed in the most developing nations. The association between oral and bone health has been well researched and accepted. The aim of the study is to develop a computer-aided system to diagnose osteoporosis from digital dental panoramic radiograph (DPR) in an economical manner. DPR was obtained from older women aged greater than 40 with an informed consent in their own language for 22 females under consideration. The DPR was taken using X MIND PANO D+ scanner and Bone mineral density (BMD) assessment at skeletal sites (forearm, heel) using p-DXA system (OsteoSys EXA 3000). The WHO classification scheme was adopted to classify the subjects as normal and osteoporotic groups. The ROI was manually marked size of $256 \times 256$ pixels around the mental foramen and the acquired DPR were preprocessed by means of contrast stretching, histogram equalization thresholding and high pass filtering to attain the foreground information. The cortical bone boundaries were determined by means of active contour. The distance between the lower and the upper border of mandible was determined by Euclidean distance approach and designated as Mandibular cortical width (MCW). It was found that the MCW measured using the proposed system could exhibit appreciable results in comparison with the measurement made with the MCW determined using the manual approach. The results of the study infer that MCW measured from DPR could be useful in diagnosing Osteoporosis.