Automated speech analysis demonstrates utility in Huntington’s disease

  • A study of 69 people with Huntington’s disease (HD; 44 premanifest and 25 manifest) and 25 matched healthy controls used quantitative automated acoustic analysis to evaluate patterns of speech alteration in 10 dimensions.
  • The automated analysis differentiated between people with and without HD, even in the early stages of the disease, with an accuracy of 74%, 92%, and 97% for the presymptomatic, prodromal, and manifest stages, respectively.
  • Speech abnormalities also correlated with motor and cognitive deficits, leading the researchers to conclude that automated speech analysis could represent a quantitative biomarker to assess HD progression.