AUTHOR=Cui Jingnan , Liu Cheng Lei , Jennane Rachid , Ai Songtao , Dai Kerong , Tsai Tsung-Yuan 

TITLE=A highly generalized classifier for osteoporosis radiography based on multiscale fractal, lacunarity, and entropy distributions

JOURNAL=Frontiers in Bioengineering and Biotechnology

VOLUME=Volume 11 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/bioengineering-and-biotechnology/articles/10.3389/fbioe.2023.1054991

DOI=10.3389/fbioe.2023.1054991

ISSN=2296-4185

ABSTRACT=Background: Osteoporosis is a common and high-incidence degenerative disease for aging populations. However, asymptomatic osteoporosis is often overlooked or misdiagnosed. A typical annual physical examination usually involves radiography but does not include tests for bone mineral density (BMD) or bone trabecular condition. Therefore, we proposed a highly generalized classifier for osteoporosis radiography based on the multi-scale fractal, lacunarity, and entropy distributions.
Methods: We collected 104 radiographs (92 for training, 12 for testing) of lumbar spine L4 and divided them into three classes (normal, osteopenia, and osteoporosis). Additionally, 174 radiographs (116 for training, 58 for testing) of calcaneus from health and osteoporotic fracture groups were collected. The texture features of all the radiographs were extracted and analyzed. Davies Boulding Index (DBI) was employed to optimize hyperparameters of feature counting. Neighbourhood Component Analysis (NCA) was applied to reduce feature dimension and increase generalization. An SVM classifier was trained with only the most effective 6 features for each binary classification scenario. The performance was estimated by the area under the curve (AUC), accuracy and sensitivity.
Results: Interpretable feature trends of osteoporotic pathological changes were depicted. On spine test dataset, AUC, accuracy and sensitivity of binary classifiers were respectively 0.851 (95% CI: 0.730-0.922), 0.813 (95% CI: 0.718-0.878) and 0.936 (95% CI: 0.826-1) for osteoporosis diagnose; 0.721 (95% CI: 0.578-0.824), 0.675 (95% CI: 0.563-0.772) and 0.774 (95% CI: 0.635-0.878) for osteopenia diagnose; and 0.935 (95% CI: 0.830-0.968), 0.928 (95% CI: 0.863-0.963) and 0.910 (95% CI: 0.746-1) for osteoporosis diagnose from osteopenia. On calcaneus test dataset, 0.767 (95% CI: 0.629-0.879), 0.672 (95% CI: 0.545-0.793) and 0.790 (95% CI: 0.621-0.923) for osteoporosis diagnose.