Johannes Tröger, Felix Dörr, Louisa Schwed, Nicklas Linz, Alexandra König, Tabea Thies, Tereza Tykalova, Michal Novotny, Petr Dusek, Jan Rusz
*Poster presented at ADPD 2025
Abstract
Aims: Isolated REM sleep behavior disorder (iRBD) is an early alpha-synucleinopathy and can precede overt motor Parkinson’s Disease (PD) for decades. Speech changes have proven to be a robust and sensitive indicator of motor-function impairment in PD. Recently automatic speech analysis has shown to be an early biomarker of motor dysfunction in individuals with iRBD (Hlavnička et al., 2017). The goal of this research is to evaluate the feasibility of using speech biomarkers to screen iRBD patients.
Methods: This research is based on Czech data with 60 iRBD (11F, mean age = 67.08 ± 7.42, mean UPDRS III = 6.93 ± 4.88) and 60 HC (11F, mean age = 65.18 ± 7.35, mean UPDRS III = 4.8 ± 3.96) (Rusz et al., 2021) with recordings from fast syllable repetitions of the sequence /pataka/. From the recordings, acoustic features on diadochokinetic rate and irregularity were extracted using ki:elements’ proprietary speech analysis pipeline SIGMA. Acoustic features were z-standardized. Then, we calculated a composite score for articulatory performance, computed Kruskal-Wallis group difference between HC and iRBD and subsequently defined an optimal cut-off to separate both groups and report the area under the curve for this screening use case. Each participant was also assessed with the Unified Parkinson’s Disease Rating Scale (UPDRS).
Results: The articulation score significantly differs between both groups (H = 22.663, p < 0.001, η2 = 0.184, Cohen’s d = 0.472) with the iRBD group showing significantly less articulatory performance in the /pataka/ task. Using the overall articulation score and a cut-off, iRBD patients could be screened out with an AUC of .75 (sensitivity = 0.80, specificity = .65). The receiver operating characteristics can be seen in the figure below.
Conclusions: The results show that an automatic articulation measure based on acoustic speech biomarkers could be utilized to screen for iRBD. Future research has to show whether this result persists in a study with different languages. Results might impact future applications both in clinical trials as well as healthcare.
