Identification of abnormal knee joint vibroarthrographic signals based on fluctuation features

In this work, we extracted the variation features and performed the pattern classifications for knee joint vibroarthro-graphic (VAG) signal processing. The signal turns count with fixed threshold and coefficient of variation (CV) of envelope energy were used to characterize the intrinsic oscillations of the VAG signals. The Kolmogorov-Smirnov test results indicated the pathological VAG signals possess significantly different signal turns count with fixed threshold and CV of envelope energy values (p < 0.01) from the healthy normal signals. The classification experiment results demonstrated that the Bayesian decision rule can produce an overall classification accuracy of 84%, with a sensitivity value of 0.75 and a specificity value of 0.894.