Pellet placement in x-ray microbeam studies

A weakness of the x-ray microbeam system for recording articulatory movements is the difficulty of placing pellets on the pharyngeal part of the tongue. It is, however, possible to use the system to obtain static position of the tongue-root in another way. Image scans of sustained English vowels of one subject, in which a chain of 15 pellets 5 mm apart had been placed on the midline of the tongue, were recorded and analyzed. The results enabled the determination of the best locations for placing pellets and the extent to which tongue shapes can be determined from a small number of pellets. If only two pellets are used, one should be placed in the range of 30–35 mm and the other 60–65 mm from the tip. The tongue shapes predicted from pellets in these locations were highly correlated (r2 = >0.98) with the observed shapes as specified by three factors that themselves determine the tongue shapes within 0.5 ram. The same high correlations (r2 = >0.98) can be achieved with several different combinations of three pellets. Most (75%) of these combinations involve one point at the root of the tongue. However, by selecting points near the tongue tip (5–10 mm back), the middle (approximately 35 mm from the tip) and the back (45–50 mm from the tip), equally good correlations are obtained.