Analytical validation of digital measure of intelligibility in progressive supranuclear palsy

İlkay Yıldız Potter; Ram Kinker Mishra; Tejas Pawar; Jesse Wang; Antje S. Mefferd; Michael de Riesthal; Mili Kuruvilla-Dugdale; Anne-Marie Wills; Alexander Pantelyat; Ashkan Vaziri

doi:10.3389/fnins.2026.1961535

Back

Analytical validation of digital measure of intelligibility in progressive supranuclear palsy

Journal article

Open access

Peer reviewed

Analytical validation of digital measure of intelligibility in progressive supranuclear palsy

İlkay Yıldız Potter, Ram Kinker Mishra, Tejas Pawar, Jesse Wang, Antje S. Mefferd, Michael de Riesthal, Mili Kuruvilla-Dugdale, Anne-Marie Wills, Alexander Pantelyat and Ashkan Vaziri

Frontiers in neuroscience, Vol.20, 1961535

09/28/2026

DOI: 10.3389/fnins.2026.1961535

PMCID: PMC13655066

PMID: 42865502

Files and links (1)

url

https://doi.org/10.3389/fnins.2026.1961535View

Published (Version of record) Open Access

Abstract

Introduction: Progressive supranuclear palsy (PSP) causes progressive speech impairment. Digital speech analysis can objectively and remotely quantify measures like intelligibility, typically judged subjectively by experts during clinical visits. The objective of this study is to validate digital intelligibility measurement in PSP via BioDigit Speech (BioSensics LLC, United States). Methods: Forty-nine PSP patients performed standardized passage-reading scored by BioDigit Speech and three experts (18 + years of experience); expert mean score (EMS) served as the reference standard. Results: Digital intelligibility showed high correlation and agreement with EMS (r = 0.90, ICC = 0.84 in native speakers), while systematically scoring below EMS. Moreover, the digital measure discriminated PSP clinical severity as well as or better than EMS, particularly for PSP-QoL-11 with a steeper standardized response. Correlations of digital measure with the clinical scales were moderate (up to −0.51), significant (p-values <0.02) and consistently higher than the ones of EMS. Raters appeared to use top-down information to recover underlying messages, resulting in saturated scores near 1.0 and low-to-moderate inter-rater agreement. Conclusion: Collected remotely with minimal burden, digital speech measures like intelligibility may serve as objective, sensitive outcomes for communication-focused interventions and PSP trials.

digital health

dysarthria

intelligibility

progressive supranuclear palsy

speech analysis and processing

Details

Title: Subtitle: Analytical validation of digital measure of intelligibility in progressive supranuclear palsy
Creators: İlkay Yıldız Potter
Ram Kinker Mishra - BioSensics (United States)
Tejas Pawar - BioSensics (United States)
Jesse Wang - Harvard University
Antje S. Mefferd - Vanderbilt University Medical Center
Michael de Riesthal - Vanderbilt University Medical Center
Mili Kuruvilla-Dugdale - University of Iowa
Anne-Marie Wills - Harvard University
Alexander Pantelyat - Johns Hopkins University
Ashkan Vaziri - BioSensics (United States)
Resource Type: Journal article
Publication Details: Frontiers in neuroscience, Vol.20, 1961535
DOI: 10.3389/fnins.2026.1961535
PMID: 42865502
PMCID: PMC13655066
ISSN: 1662-453X
eISSN: 1662-453X
Publisher: Frontiers Media SA
Grant note: National Institute on Aging of the National Institutes of Health: R44AG080861, UF1AG094587
The author(s) declared that financial support was received for this work and/or its publication. Research reported in this publication was supported by the National Institute on Aging of the National Institutes of Health under Award Numbers R44AG080861 and UF1AG094587. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Language: English
Date published: 09/28/2026
Academic Unit: Neurology; Communication Sciences and Disorders
Record Identifier: 9985240904402771

Metrics

10 Record Views