Classification of Manifest Huntington Disease using Vowel Distortion Measures

Predicting clinical scores in Huntington's disease: a lightweight speech test

OBJECTIVES

Using brief samples of speech recordings, we aimed at predicting, through machine learning, the clinical performance in Huntington's Disease (HD), an inherited Neurodegenerative disease (NDD).

METHODS

We collected and analyzed 126 samples of audio recordings of both forward and backward counting from 103 Huntington's disease gene carriers [87 manifest and 16 premanifest; mean age 50.6 (SD 11.2), range (27-88) years] from three multicenter prospective studies in France and Belgium (MIG-HD (ClinicalTrials.gov NCT00190450); BIO-HD (ClinicalTrials.gov NCT00190450) and Repair-HD (ClinicalTrials.gov NCT00190450). We pre-registered all of our methods before running any analyses, in order to avoid inflated results. We automatically extracted 60 speech features from blindly annotated samples. We used machine learning models to combine multiple speech features in order to make predictions at individual levels of the clinical markers. We trained machine learning models on 86% of the samples, the remaining 14% constituted the independent test set. We combined speech features with demographics variables (age, sex, CAG repeats, and burden score) to predict cognitive, motor, and functional scores of the Unified Huntington's disease rating scale. We provided correlation between speech variables and striatal volumes.

RESULTS

Speech features combined with demographics allowed the prediction of the individual cognitive, motor, and functional scores with a relative error from 12.7 to 20.0% which is better than predictions using demographics and genetic information. Both mean and standard deviation of pause durations during backward recitation and clinical scores correlated with striatal atrophy (Spearman 0.6 and 0.5-0.6, respectively).

INTERPRETATION

Brief and examiner-free speech recording and analysis may become in the future an efficient method for remote evaluation of the individual condition in HD and likely in other NDD.

Feasibility of a continuous, multi-sensor remote health monitoring approach in persons living with neurodegenerative disease

BACKGROUND

Remote health monitoring with wearable sensor technology may positively impact patient self-management and clinical care. In individuals with complex health conditions, multi-sensor wear may yield meaningful information about health-related behaviors. Despite available technology, feasibility of device-wearing in daily life has received little attention in persons with physical or cognitive limitations. This mixed methods study assessed the feasibility of continuous, multi-sensor wear in persons with cerebrovascular (CVD) or neurodegenerative disease (NDD).

METHODS

Thirty-nine participants with CVD, Alzheimer's disease/amnestic mild cognitive impairment, frontotemporal dementia, Parkinson's disease, or amyotrophic lateral sclerosis (median age 68 (45-83) years, 36% female) wore five devices (bilateral ankles and wrists, chest) continuously for a 7-day period. Adherence to device wearing was quantified by examining volume and pattern of device removal (non-wear). A thematic analysis of semi-structured de-brief interviews with participants and study partners was used to examine user acceptance.

RESULTS

Adherence to multi-sensor wear, defined as a minimum of three devices worn concurrently, was high (median 98.2% of the study period). Non-wear rates were low across all sensor locations (median 17-22 min/day), with significant differences between some locations (p = 0.006). Multi-sensor non-wear was higher for daytime versus nighttime wear (p < 0.001) and there was a small but significant increase in non-wear over the collection period (p = 0.04). Feedback from de-brief interviews suggested that multi-sensor wear was generally well accepted by both participants and study partners.

CONCLUSION

A continuous, multi-sensor remote health monitoring approach is feasible in a cohort of persons with CVD or NDD.