The acoustical analysis of knee joint sounds for non-invasive diagnosis of articular pathology

We have evaluated and classified arthritic pathology using the acoustical analysis of knee joint sounds. Six normal subjects and 11 patients with knee problem were enrolled and patients were divided into the 1st patient group needed an orthopaedic surgery because of the ruptured wounds of meniscus or ACL (anterior cruciate ligament) and 2nd patient group diagnosed as osteoarthritis. In sitting and standing position, subjects flexed and extended the knee joint, the sounds and angle of knee joint were collected. Dynamic time warping was applied for normalizing a time-axis. The fundamental frequency, the mean amplitude of pitch, jitter, shimmer were analyzed according to a position and unpaired T-test and ANOVA test were used for statistic results. The experimental results showed that the fundamental frequency of the 2nd patient group or standing groups was higher than the others and the pitch of sounds was varied unstably.