Back

Compensatory mechanisms enables intelligibility in prodromal Parkinson's disease

Thies, T.; Muecke, D.; Geerts, N.; Seger, A.; Fink, G. R.; Barbe, M. T.; Sommerauer, M.

2022-12-14 neurology
10.1101/2022.12.14.22283451 medRxiv
Show abstract

BackgroundSpeech impairment is already present on the acoustic level in speakers with isolated rapid eye movement sleep behavior disorder (iRBD). The aim of this study was to determine whether speech changes are already present on the articulatory level and if how these differ from healthy control speakers and speakers with Parkinsons disease (PD). MethodsKinematic data were collected from 68 age and sex-matched subjects: healthy control speakers (n=23), patients with iRBD (n=22), and patients with PD (n=23). All participants were recorded with electromagnetic articulography (AG 501) to capture articulatory movements of the lower lip, the tongue tip and the tongue body. Movement amplitudes, durations and average speeds were calculated per articulator. In addition, naive listeners rated the intelligibility of the speech sampled produced by the participants. ResultsThe results of the production experiment indicate changes between the control and the iRBD group as well as between the iRBD and the PD group. Movement durations increase from the control group to the iRBD group and further to the PD group. In contrary, movement amplitudes increase in prodromal PD and decreased from PD onwards. This relationship was reflected by preserved articulatory speed in iRBD patients, but articulatory slowdown in PD. Intelligibility is lower in speakers with PD, but does not differ between the control and the iRBD group. ConclusionSpeakers with iRBD adjust underlying articulatory movement patterns to maintain their intelligibility level.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.