The loss of bone mineral density (BMD) in various sites of the body is the main sign of osteoporosis. Osteoporosis is referred to as a “silent disease” because it does not cause obvious symptoms and is usually detected when the bone fracture has already presented. BMD measurement using a dual-energy X-ray absorptiometry (DEXA) device is the gold standard for diagnosing osteoporosis by assessing the condition of the bones. Since low BMD can affect any skeletal bone, osteoporosis can also manifest in the mandible. Many studies have reported that digital panoramic radiographs can be used as a tool to assist in the detection of osteoporosis by measuring and analyzing the cortical bone width. Early detection of osteoporosis is essential to prevent more damage. This prevention can be done by detecting the early stage of bone loss by low BMD, which is commonly referred to as osteopenia. Women with mild osteopenia can develop osteoporosis in 15 years, while women with advanced osteopenia can develop osteoporosis within a year. Therefore, lifestyle modification in osteopenia can be done to prevent bone mass and structural deterioration leading to osteoporosis. Although osteopenia has a lower risk of fracture compared with osteoporosis, its detection becomes more important because the prevalence of fracture in postmenopausal women with osteopenia is 21%. Moreover, the number of persons with osteopenia is much higher than the number of persons with osteoporosis. Therefore, a substantial proportion of the population is at risk of fracture.
Digital panoramic radiographs have been used in several radiomorphometric studies to predict osteoporosis based on the BMD level. Mandibular cortical width (MCW), panoramic mandibular index, and Klemetti index are overall useful tools that potentially could be used by dentists to screen for low BMD and predict osteoporosis. Since the panoramic radiograph is commonly used as a diagnostic tool in dental practice, it will be very useful for the initial screening of the osteoporosis risk and recommending the examination of additional clinical risk factors to prevent osteoporotic fracture. However, manually measuring the potential parameters in a panoramic radiograph is time-consuming and prone to measurement errors.
To assist clinicians, automated osteoporosis detection using digital panoramic radiographs is generally performed by measuring the width of cortical bone at the mandible. A computer-aided system for identifying osteoporosis using a digital panoramic radiograph revealed that the correlation between cortical bone width and BMD measurement result is approximately 53%. In this accord, the study by Geary et al also demonstrated a significant correlation between the width of cortical bone and bone density, but the significance level decreased when combined with other factors, such as age, length, and width of the jaw, and the number of teeth, to detect bone density. Several studies have been conducted to detect osteoporosis by measuring the bone density of the mandible using cone-beam computed tomography (CBCT) imaging.
In this study, we developed a computer-aided diagnosis (CAD) system to assist osteoporosis assessment based not only on MCW but also on analysis of the texture of the cortical bone representing the bone porosity. This assessment was performed using digital panoramic radiographs and patients were classified into normal, osteopenia, or osteoporosis groups using multiclass classification with the support vector machine (SVM). To validate the proposed CAD system, we compared the classification result of the CAD system to the BMD measurement of the patient using the DEXA device. The experimental results using 10-fold cross-validation showed that the proposed system achieved an average accuracy of 86.50% for osteoporosis classification on dental panoramic radiographs. The average misclassification error and relative foreground area error of the segmentation process were 5.21%. The assessment of the confusion matrix was done by two dental radiologists (E.R.A. and B.P.). The dental radiologist determined the ROI for the osteoporosis assessment in the system and manual osteoporosis assessment. Intraobserver agreement of the osteoporosis assessment by using CAD was 74 and 78% for the first and second dental radiologists, respectively. Interobserver agreement of the osteoporosis assessment by using CAD between the two radiologists was 81%.
In this research, a CAD system was proposed to measure width or thickness and analyze cortical bone texture in the mandibular cortical bone in panoramic dental images. This system can be used to assist in osteopenia and osteoporosis detection, with an average system of more than 80%. The use of cortical bone width and texture features can provide a higher system accuracy value than using only the cortical bone width feature. Further studies compared with other types of radiographic imaging such as CBCT may be needed so that radiographic examination can be one of the methods for early detection of osteoporosis accurately.
Author: Eha Renwi Astuti
View more: https://www.thieme-connect.com/products/ejournals/html/10.1055/s-0042-1749158
