Can a snore recording distinguish palatal from non-palatal snoring?

Details

Distinguishing palatal from non-palatal snoring is important for ENT surgeons deciding whether palatal surgery might be curative. Sleep nasendoscopy is currently the accepted method.

Palatal vibration produces marked changes in sound loudness at low frequencies, below approximately 100 Hz. To measure the degree of modulation, the researchers calculated the crest factor of the sound waveform: the ratio of its peak value to its root mean square (rms) value in a given epoch.

The researchers used a digital tape recorder to record free-field snore sounds from 11 adult patients lying on their backs under intravenous midazolam sedation. The recordings were transferred to a PC at a sampling frequency of 11 kHz and analysed with code written by the researchers. Simultaneous sleep nasendoscopy provided direct visual confirmation of the snoring site and was taken as the gold standard.

The dominant site was the soft palate in six patients. The five non-palatal cases comprised one epiglottic, two hypopharyngeal and two tongue-base snorers. The crest factor was significantly higher in palatal snorers (p < 0.01, Student-t or Mann-Whitney tests). Crest factor alone distinguished palatal from non-palatal snoring in all 11 cases, making it a promising non-invasive diagnostic technique.