Detection of voice changes due to aspiration via acoustic voice analysis

Aspiration can occur during swallowing. The present study investigated the feasibility of identifying aspiration risk (AR) via acoustic voice parameters.

In total, 165 patients scheduled for a videofluoroscopic swallowing study (VFSS) were included. The patients were divided into two groups (AR and non-AR) based on their VFSS results. The AR group, which had ingested materials on or below the vocal folds, included 59 patients (42 males and 17 females). The non-AR group, which showed normal swallowing, included 106 patients (49 males and 57 females). The major cause of swallowing disorders was a stroke. A sustained vowel/a/for at least 3s was recorded before and after swallowing. Eight acoustic voice parameters were measured using PRAAT, including fundamental frequency, standard deviation of F0, jitter, relative average perturbation (RAP), shimmer, amplitude perturbation quotient (APQ), harmonic-to-noise ratio (HNR), and noise-to-harmonic ratio (NHR). Changes in each acoustic voice parameter before and after the VFSS were compared between the two groups with a repeated-measures mixed analysis of variance.

Only RAP showed a statistically significant interaction between group (non-AR and AR) and time (pre- and post-swallowing; p=0.030). RAP decreased after swallowing in the AR group; however, it increased in the non-AR group. Jitter and NHR increased in the non-AR group but decreased in the AR group after swallowing, but the difference was not statistically significant.

Our results suggest that the accumulation of pasty food in the vocal folds may modify vocal fold vibration and change voice quality in patients with penetration/aspiration. Several acoustic voice parameters, especially jitter, RAP, and NHR, were affected by AR. Thus, acoustic voice analysis may be helpful in making a diagnosis of AR as a supplementary tool for standard swallowing study including VFSS or fiberoptic examination.