Osteoporosis Diagnosis Based on the Multifractal Spectrum Features of Micro-CT Images and C4.5 Decision Tree
Zhi Yang Gao, Wenxue Hong, Yonghong Xu, Tao Zhang, Zhijie Song, Jian Liu · 2010
In this work the multifractal spectrum features of micro-CT images were extracted and classified the features of multifractal spectrum using C4.5 decision trees which can help doctors to diagnose osteoporosis. Micro-computed tomography (micro-CT) allows for classical anatomical imaging well suited to the study of skeletal structures, and the multifractal spectrum can reflect the complex and self-similarity characteristic of structures. So we extracted the multifractal spectrum features of trabecular bone micro-CT images, and used C4.5 to classify. The results shown the proposed method has a higher accuracy, and this method adopted here could be a useful supplement to the clinical observations for osteoporosis diagnosis.