Osteoarthritis detection system using optimal dynamic feature configuration

Cepstral analysis has been invariably applied in processing various signals in which include vibroarthrography (VAG). The cepstral analysis, however, requires substantial efforts to optimize its performance on VAG signal before real application for osteoarthritis detection is possible. In this paper, therefore, we attempt to analyze the VAG signals by implementing delta and energy features. The analysis involves 30 volunteers and 900 acquired VAG signals. Our study shows that delta and energy features are capable to signalize osteoarthritis to a certain extent with superior computational feasibility. In conclusion, this paper explores the applicability of energy property in analyzing the VAG signal for future VAG study. Furthermore, via this paper, we intent to draw attention of future research, particularly on the correlation among various clinical parameters, pathologic VAG signals and muscle contrary interference (MCI) effect.