Abstract

Electromagnetic articulography (EMA) have been mostly employed to study articulatory movements of speech. This technology appears to be very promising for studying mandibular movements within the field of dentistry. However, there are no studies reporting the validity of EMA for such purpose. The aim of this study is to assess accuracy and reliability of Carstens three-dimensional EMA AG501 in order to validate its use for mandibular movement analysis in dentistry. A set of tests was conducted attaching 16 sensors to a rotating rigid structure placed inside the measurement area. Another set of tests were conducted using a mouth anatomical model with human-like articulatory behaviour. A function of the EMA system called "head correction" was applied to normalize the data of every recording. The system reliability was higher at the centre of the measurement area and decreased toward the edges. Dispersion was greater for raw data than for normalized data. Bland-Altman analysis of agreement between the AG501 and a millimetre ruler used to measure the distance between the sensors revealed limits of agreement between 0.5 mm and -0.9 mm. The results suggest that EMA AG501 is valid for three-dimensional analysis of mandibular biomechanics allowing natural movements.