Abstract
The use of a microphone to detect sound emitted from knee joint and tissues is a significant advance on acoustic detection systems. All possible artifact signals encountered by the sound detection technique were investigated. 1) Background noise is minimized by measuring in an anechoic chamber. 2) Coupler noise is eliminated by ventilation holes. 3) Skin friction noise is effectively eliminated using ultrasonic-gel. 4) The use of a double microphone with differential amplifier to subtract the signals between two microphones is not suitable for the measurement of knee joint sound, because of the appearance of peak and dip frequency in a spectrum. 5) It is possible to obtain the actual spectrum of knee joint sound by subtracting the spectrum of background sound from the spectrum of recorded knee joint sound.
