1School of Dentistry and 2Imaging Science and Biomedical Engineering, The University of Manchester, United Kingdom 3School of Dentistry, Oral Pathology and Maxillofacial Surgery, Oral Imaging Centre, Katholieke Universiteit Leuven, Belgium 4Oral Diagnosis and Radiology, The National and Kapodistrian University of Athens, Greece 5Oral Radiology, Malmö University, Sweden 6Oral Radiology, ACTA, Amsterdam, The Netherlands Objectives: To determine the diagnostic validity of the width of the inferior mandibular cortex on dental panoramic radiographs (DPRs), as measured by Active Shape Model (ASM) methods, for the diagnosis of osteoporosis in peri- and post-menopausal women. Methods: Volunteer female subjects in the 45 to 70 year age band, recruited from four European centres, underwent dual x-ray energy absorptiometry of the hip and lumbar spine, to provide a gold standard diagnosis of osteoporosis, and a DPR examination. Two ASM methods, one entirely automatic and one manually initialised, were used to derive measurements of mandibular cortical width. ROC analysis was used to determine the diagnostic ability of manual and ASM methods. Results: 652 subjects were examined (mean age 54.9y; sd=6.10), with 140 (21.5%) being classified as having osteoporosis. Using the manually initialized ASM method, Az values for the diagnosis of osteoporosis at any site and at femoral neck alone were 0.816 (95% CI: 0.784 to 0.845) and 0.835, (95% CI: 0.805-0.863), respectively. Using the automatically initialized ASM method, the Az values for the diagnosis of osteoporosis at any site and at the femoral neck alone were 0.759 (95% CI: 0.724-0.791) and 0.805 (95% CI: 0.773 -0.835), respectively. The difference in Az of the two methods for the diagnosis of osteoporosis at either the hip or spine was significant (p<0.001), but not significant at the femoral neck alone. Conclusions: ASM-based methods ofmandibular cortical width measurement were effective in the diagnosis of osteoporosis. The manually initialized method, involving a small amount of user interaction, performed best. Further analysis is necessary to establish the appropriate diagnostic threshold for clinical use. This work was supported by a research and technological development project grant from the European Commission FP5 'Quality of Life and Management of Living Resources' (QLK6-2002-02243).