Mandibular Trabecular Bone Analysis Using Local Binary Pattern for Osteoporosis Diagnosis
Background: Osteoporosis is a systemic skeletal disease characterized by low bone mineral density (BMD) and micro-architectural deterioration of bone tissue, leading to bone fragility and increased fracture risk. Since Panoramic image is a feasible and relatively routine imaging technique in dent...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Shiraz University of Medical Sciences
2019-02-01
|
| Series: | Journal of Biomedical Physics and Engineering |
| Subjects: | |
| Online Access: | http://jbpe.ir/Journal_OJS/JBPE/index.php/jbpe/article/view/743/459 |
| Summary: | Background: Osteoporosis is a systemic skeletal disease characterized by low
bone mineral density (BMD) and micro-architectural deterioration of bone tissue,
leading to bone fragility and increased fracture risk. Since Panoramic image is a
feasible and relatively routine imaging technique in dentistry; it could provide an
opportunistic chance for screening osteoporosis. In this regard, numerous panoramic
derived indices have been developed and suggested for osteoporosis screening. Jaw
trabecular pattern is one of the main bone strength factors and trabecular bone pattern
assessment is important factor in bone quality analysis. Texture analysis applied to
trabecular bone images offers an ability to exploit the information present on conventional radiographs.
Objective: The purpose of this study was to evaluate the relationship between
Jaw trabecular pattern in panoramic image and osteoporosis based on image texture
analyzing using local binary pattern.
Material and Methods: An experiment is evaluated in this paper based on a
real hand-captured database of panoramic radiograph images from osteoporosis and
non-osteoporosis person in Namazi Hospital, Shiraz, Iran .An approach is proposed
for osteoporosis diagnosis consisting of two steps. First, modified version of local
binary patterns is used to extract discriminative features from jaw panoramic radiograph images. Then, classification is done using different classifiers.
Results: Comparative results show that the proposed approach provides classification accuracy about 99.6%, which is higher than many state-of-the-art methods.
Conclusion: High classification accuracy, low computational complexity, multiresolution and rotation invariant are among advantages of our proposed approach. |
|---|---|
| ISSN: | 2251-7200 2251-7200 |