Osteoporosis Detection Using Combined Texture Features of Proximal Femur Radiographs | ||||
Mansoura Journal for Computer and Information Sciences | ||||
Volume 15, Issue 2, December 2019, Page 27-34 PDF (1.21 MB) | ||||
Document Type: Original Research Articles. | ||||
DOI: 10.21608/mjcis.2019.321065 | ||||
View on SCiNiTO | ||||
Authors | ||||
Heba Khaled* ; Nagham Mekky; Ahmed Atwan; Hassen Soliman | ||||
Faculty of Computers and Information, Information Technology Mansoura University, Egypt | ||||
Abstract | ||||
This paper presents a computer-aided-detection system of osteoporosis. The proposed technique is implemented and applied to 79 proximal femur radiographs. A dual-energy xray absorptiometry (DEXA) scan is used to measure T-score of the images as a justification. Three feature extraction techniques are introduced to describe trabecular pattern changes in proximal femur recorded: wavelet-based hierarchical pyramid, Gabor filter, and intensity gradient map. The selected features were utilized in the design and training of support vector machine (SVM) classifier. The accuracy, sensitivity, and specificity are used to measure the quality of the proposed detection system. The best result and detect femur bone fractures and osteoporosis were obtained efficiently by using wavelet-based hierarchical approach combined with Gabor filter, and intensity gradient map features. The proposed system showed superior performance as compared to other related work. | ||||
Keywords | ||||
Dual-energy x-ray absorptiometry (DEXA); Bone mineral density (BMD); Feature extraction; Osteoporosis; X-ray imaging | ||||
Statistics Article View: 19 PDF Download: 17 |
||||