Digital App for Speech & Health Monitoring

Not yet recruiting

• Conditions: Dementia; Multiple Sclerosis; Motor Neuron Disease; Parkinson Disease

• Registration Number: NCT06450418
• Lead Sponsor: University of Edinburgh

Brief Summary

Many people living with neurodegenerative conditions like dementia, motor neuron disease (MND), multiple sclerosis (MS), and Parkinson's disease (PD), suffer from speech problems. Using common digital technologies such as smartphone apps, the investigators can record and analyse speech in detail to provide new information for people living with these conditions, researchers, and healthcare professionals. This study will investigate the use of these digital speech recordings to help diagnose and monitor these conditions.

To take part, participants will have either a diagnosis of dementia, motor neuron disease, Parkinson's disease or Multiple Sclerosis, OR they will have no diagnosis of a neurological condition. Researchers will compare people with a diagnosis of a Neurological condition to those without.

Detailed Description

This project aims to create novel speech-based solutions for: 1) Early detection, 2) Monitoring and 3) Stratification of neurodegenerative disorders including dementia, motor neuron disease (MND), Parkinson's disease (PD), and multiple sclerosis(MS). The investigators will develop and validate proof of concept and early-stage algorithms derived from acoustic data, which will be scaled and tested in deeply-phenotyped population.

2.2 Objectives Primary Objectives

1. To deploy and iterate a digital platform, co-produced with people living with neurodegenerative disorders, for acquisition of speech data from well characterised cohorts of people living with neurodegenerative disorders (dementia, motor neuron disease, multiple sclerosis, Parkinson's disease), and a healthy control cohort (comprising relatives/carers and volunteers without a neurological diagnosis), linked to our highly curated clinical registries at the Anne Rowling Regenerative Neurology Clinic.

2. To collect a large body of acoustic speech data from well characterised cohorts of people living with neurodegenerative disorders (dementia, MND/ALS, multiple sclerosis, Parkinson's disease), and a healthy control cohort (comprising relatives/carers and volunteers without a neurological diagnosis), linked to highly curated clinical registries.

3. To apply machine learning approaches directly to acoustic and linguistic signals from voices from people with dementia, MND, MS, Parkinson's, and healthy controls (comprising relatives/carers and volunteers without a neurological diagnosis), and to characterise prosodic patterns (rhythm, intonation, and fluency) without explicit reference to the text which is spoken, providing powerful cues about the health of the speaker.

4. Compare speech based digital outcome measures to current clinical standards to characterise and validate their clinimetric properties.

Secondary Objectives

1. Assess the feasibility and acceptability of a digital outcome measure platform in people living with neurodegenerative conditions, for use in clinical care and research.

2. To create a repository of well characterised acoustic voice samples for open access sharing/collaboration with research and industry partners.

Study Timeline

• Status Verified: 2024-01
• Study First Submitted: 2024-02-09
• Study Start: 2024-06
• Study First Submitted QC: 2024-06-03
• Last Update Submitted: 2024-06-03
• Study First Posted: 2024-06-10
• Last Update Posted: 2024-06-10
• Primary Completion: 2026-06
• Study Completion: 2026-06

Recruitment & Eligibility

• Status: NOT_YET_RECRUITING
• Sex: All
• Target Recruitment: 150

Inclusion Criteria

Not provided

Exclusion Criteria

• Age <16 years
• Significant and uncorrected visual or hearing impairment (precluding use of the App).
• Lack capacity to consent to project due to cognitive impairment (precluding understanding of the study and use of the App).

Study & Design

• Study Type: OBSERVATIONAL
• Study Design: Not specified

Primary Outcome Measures

Name	Time	Method
Primary outcome measures	24 months	Area under the curve (AUC) of the receiver operating characteristic (ROC) curve for each of the 4 binary classifiers distinguishing between a disease-positive group and a healthy control group.

Secondary Outcome Measures

Name	Time	Method
Secondary outcome measure	24 months	Mean squared error of 4 regression models making predictions of condition-specific clinical rating scores (ACE-III, ALSFRS-R, EDSS, MDS-UPDRS)

Trial Locations

Locations (1)

NHS Lothian; 🇬🇧 Edinburgh, United Kingdom