The feasibility of using acoustic markers of speech for optimizing patient outcomes during randomized amplitude variation in deep brain stimulation: A proof of principle methods study

Abstract

Background: Deep brain stimulation (DBS) is an effective treatment for reducing symptoms of tremor. A common and typically subjectively determined adverse effect of DBS is dysarthria. Current assessment protocols are driven by the qualitative judgments of treating clinicians and lack the sensitivity and objectivity required to optimize patient outcomes where multiple stimulation parameters are trialed. Objective: To examine the effect of DBS on speech in patients receiving stimulation to the posterior sub-thalamic area (PSA) via randomized manipulation of amplitude parameters. Methods: Six patients diagnosed with tremor receiving treatment via DBS of the PSA were assessed in a double-blinded, within-subjects experimental protocol. Amplitude (i.e., voltage or current) was randomly adjusted across 10 settings, while speech samples (e.g., sustained vowel, counting to 10) were recorded to identify the patient-specific settings required for optimal therapeutic benefit (reduced tremor) with minimal adverse effects (altered speech). Speech production between stimulation parameters was quantified using acoustic analysis. Results: Speech changed as a response to DBS but those changes were not uniform across patients nor were they generally in line with changes in amplitude with the exception of reduced vocal control and increased mean silence length in two patients. Speech outcomes did not correlate with changes in tremor. Conclusion: Intra-individual changes in speech were detected as a response to modified amplitude; however, no clear pattern was observed across patients as a group. The use of objective acoustic measures allows for quantification of speech changes during DBS optimization protocols, even when those changes are subtle and potentially difficult to detect perceptually.