Bayesian inference of state feedback control parameters for fo perturbation responses in cerebellar ataxia
Jessica L. Gaines,Kwang S. Kim,Ben Parrell,Vikram Ramanarayanan,Alvincé L. Pongos,Srikantan S. Nagarajan,John F. Houde
DOI: https://doi.org/10.1371/journal.pcbi.1011986
2024-10-12
PLoS Computational Biology
Abstract:Behavioral speech tasks have been widely used to understand the mechanisms of speech motor control in typical speakers as well as in various clinical populations. However, determining which neural functions differ between typical speakers and clinical populations based on behavioral data alone is difficult because multiple mechanisms may lead to the same behavioral differences. For example, individuals with cerebellar ataxia (CA) produce atypically large compensatory responses to pitch perturbations in their auditory feedback, compared to typical speakers, but this pattern could have many explanations. Here, computational modeling techniques were used to address this challenge. Bayesian inference was used to fit a state feedback control (SFC) model of voice fundamental frequency ( f o ) control to the behavioral pitch perturbation responses of speakers with CA and typical speakers. This fitting process resulted in estimates of posterior likelihood distributions for five model parameters (sensory feedback delays, absolute and relative levels of auditory and somatosensory feedback noise, and controller gain), which were compared between the two groups. Results suggest that the speakers with CA may proportionally weight auditory and somatosensory feedback differently from typical speakers. Specifically, the CA group showed a greater relative sensitivity to auditory feedback than the control group. There were also large group differences in the controller gain parameter, suggesting increased motor output responses to target errors in the CA group. These modeling results generate hypotheses about how CA may affect the speech motor system, which could help guide future empirical investigations in CA. This study also demonstrates the overall proof-of-principle of using this Bayesian inference approach to understand behavioral speech data in terms of interpretable parameters of speech motor control models. Cerebellar ataxia is a condition characterized by a loss of coordination in the control of muscle movements, including those required for speech, due to damage in the cerebellar region of the brain. Behavioral speech experiments have been used to understand this disorder's impact on speech motor control, but the results can be ambiguous to interpret. In this study, we fit a computational model of the neural speech motor control system to the speech data of speakers with cerebellar ataxia and the data of typical speakers to determine what differences in model parameters best explain how the two groups differ in their control of vocal pitch. We found that group differences may be explained by increased sensitivity to auditory feedback prediction errors (differences between the actual sound speakers hear of their own speech as they produce it and the sound they expected to hear) and increased motor response in individuals with cerebellar ataxia. These computational results help us understand how cerebellar ataxia impacts speech motor control, and this general approach can also be applied to study other neurological speech disorders.
biochemical research methods,mathematical & computational biology