Abstract
Acoustic analysis is regarded as an objective, reliable and non-invasive tool in voice measurement. Recently, acoustic voice analysis has become popular in clinical situations because the instrumentation is becoming more user-friendly and more affordable due to advances in digital signal processing technology. A variety of commercial acoustic analysis systems are available today. The majority of these systems have incorporated the most common perturbation measures – for example, jitter (frequency perturbation), shimmer (amplitude perturbation) and spectral noise measures – in their packages. Many investigators once considered that these three sets of objective measurements may eventually replace the subjective perceptual voice evaluation process in clinical situations. More recently, the validity of perturbation measures in analysing pathological voices has been questioned. This paper will present the findings from a study investigating the use of acoustic analysis in discriminating normal and pathological voice samples. The results showed that although jitter and shimmer could differentiate between normal and pathological voice (p < 0.05), some dysphonic subjects in the pathological group demonstrated acoustic measurements within normal limits. These results, together with findings from other published studies, suggest that jitter and shimmer measures should be used with caution in assessing pathological voice.