Validated automatic speech biomarkers in primary progressive aphasia
Open accessAbstract
Objective: To automatically extract and quantify specific disease biomarkers of prosody from the acoustic properties of speech in patients with primary progressive aphasia. Methods: We analyzed speech samples from 59 progressive aphasic patients (non‐fluent/agrammatic = 15, semantic = 21, logopenic = 23; ages 50–85 years) and 31 matched healthy controls (ages 54–89 years). Using a novel, automated speech analysis protocol, we extracted acoustic measurements of prosody, including fundamental frequency and speech and silent pause durations, and compared these between groups. We then examined their relationships with clinical tests, gray matter atrophy, and cerebrospinal fluid analytes. Results: We found a narrowed range of fundamental frequency in patients with non‐fluent/agrammatic variant aphasia (mean 3.86 ± 1.15 semitones) compared with healthy controls (6.06 ± 1.95 semitones;P < 0.001) and patients with semantic variant aphasia (6.12 ± 1.77 semitones;P = 0.001). Mean pause rate was significantly increased in the non‐fluent/agrammatic group (mean 61.4 ± 20.8 pauses per minute) and the logopenic group (58.7 ± 16.4 pauses per minute) compared to controls. In an exploratory analysis, narrowed fundamental frequency range was associated with atrophy in the left inferior frontal cortex. Cerebrospinal level of phosphorylated tau was associated with an acoustic classifier combining fundamental frequency range and pause rate (r = 0.58,P = 0.007). Receiver operating characteristic analysis with this combined classifier distinguished non‐fluent/agrammatic speakers from healthy controls (AUC = 0.94) and from semantic variant patients (AUC = 0.86). Interpretation: Restricted fundamental frequency range and increased pause rate are characteristic markers of speech in non‐fluent/agrammatic primary progressive aphasia. These can be extracted with automated speech analysis and are associated with left inferior frontal atrophy and cerebrospinal phosphorylated tau level.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Previous connected speech research has largely focused on Alzheimer disease, mild cognitive impairment, and PPA, with fewer studies focused on behavioral FTD variants.
“Previous connected speech research has largely focused on Alzheimer disease (AD), mild cognitive impairment, and PPA, with fewer studies focused on behavioral FTD variants.14-16”
From Automated Speech Analysis to Differentiate Frontal and Right Anterior Temporal Lobe Atrophy in Frontotemporal Dementia (2025) · cites this paper as [15]
The Introduction and Methods describe connected-speech analysis in PPA, and the Discussion mentions a previous acoustic analysis in behavioral-variant FTD. The paper does not establish that research has largely focused on AD, mild cognitive impairment, and PPA, or that behavioral-variant FTD has received comparatively fewer studies. (LLM, conf=0.75)
Experimental — accuracy not yet measuredHigh pause rates were found in all PPA variants, most profoundly in naPPA.
“In another study, we found high pause rates in all PPA variants, most profoundly in naPPA, 59 and in another classification study this feature differentiated subtypes of PPA. 60”
From Automatic classification of AD pathology in FTD phenotypes using natural speech (2024) · cites this paper as [59]
Across primary progressive aphasia variants, ASLA markers capture syndrome-specific patterns that predict underlying neuropathology years before death and correlate with variant-specific atrophy, even longitudinally.
“Across primary progressive aphasia variants, ASLA markers capture syndromespecific patterns 49-53 that predict underlying neuropathology years before death 7 and correlate with variant-specific atrophy, 53-56 even longitudinally. 57”
From Speech and language markers of neurodegeneration: a call for global equity (2023) · cites this paper as [53]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
The citation lists haven’t been fetched yet
References 45 · Cited by 74
Research works printed but not linked: 10
- [3]Liberman P. Intonation, perception, and language. Cambridge, MA: M.I.T Press, 1968.
- [15]Lleó A, Irwin D, Illán‐Gala I, et al. A CSF algorithm for the selection of FTLD subtypes in an autopsy cohort. JAMA Neurol 2018; 75: 738– 745. 29554190 PMC5885205
- [19]Wickham H. Reshaping data with the reshape package. J Stat Softw 2007; 21: 1– 20.
- [24]Wickham H. The split‐apply‐combine strategy for data analysis. J Stat Softw 2011; 40: 1– 29.
- [25]Venables WN, Ripley BD. Modern applied statistics with S, 4th ed New York: Springer, 2002.
- [31]Rohrer JD, Sauter D, Scott S, et al. Receptive prosody in nonfluent primary progressive aphasias. Cortex 2012; 48: 308– 316. 21047627 PMC3275751
- [36]Ladefoged P, Disner SF. Vowels and consonants. West Sussex, UK: Wiley ‐ Blackwell, 2012.
- [41]Sajjadi SA, Patterson K, Arnold RJ, et al. Primary progressive aphasia: a tale of two syndromes and the rest. Neurology 2012; 78: 1670– 1677. 22573633 PMC3359509
- [45]Irwin DJ, Lleó A, McMillan CT, et al. Ante mortem CSF tau levels correlate with post mortem tau pathology in FTLD. Ann Neurol 2017; 82: 247– 258. 28719018 PMC5776747
- [48]Boersma P. Accurate short‐term analysis of the fundamental frequency and the harmonics‐to‐noise ratio of a sampled sound. Proceedings of the Institute of Phonetic Sciences. 1993;17(University of Amsterdam):97–110.
- Authors
- Nevler, N., Ash, S., Irwin, D. J., Liberman, M., & Grossman, M.
- Affiliations
- University of Pennsylvania (US)
- Corresponding author
- Nevler, Naomi
- Grossman, Murray
- Venue
- Annals of Clinical and Translational Neurology
- Publisher
- Wiley
- Published
- 2018-11-24
- Volume
- 6
- Issue
- 1
- Pages
- 4-14
- ISSN
- 2328-9503 (ISSN-L)
- PMC
- PMC6331511
- OpenAlex
- W2900484088
- Cited by
- 74
- References
- 45
- Citations by year
- 2026: 9 · 2025: 7 · 2024: 13 · 2023: 14 · 2022: 10 · 2021: 15 · 2020: 5 · 2019: 1
- FWCI
- 4.12
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Annals of Clinical and Translational Neurology (publishedVersion, cc-by-nc-nd)
- Europe PMC (PubMed Central) (publishedVersion, other-oa)
- PubMed Central (publishedVersion, cc-by-nc-nd)
- PubMed (publishedVersion)
- Funders
- National Institutes of Health (AG017586, AG038490, AG053940, AG052943, NS088341, DC013063, AG054519)
- National Institute on Aging
- Alzheimer's Association (AACSF‐18‐567131)
- Wyncote Foundation
- Topics
- Health Sciences › Medicine › Psychiatry and Mental health › Dementia and Cognitive Impairment Research
- Health Sciences › Medicine › Physiology › Voice and Speech Disorders
- Life Sciences › Neuroscience › Cognitive Neuroscience › Neurobiology of Language and Bilingualism
- Keywords
- primary progressive aphasia, speech biomarkers, prosody, fundamental frequency range, phosphorylated tau, cerebrospinal fluid biomarkers, automated speech analysis, speech acoustics
- MeSH terms
- Aged, Aged, 80 and over, Brain, Female, Humans, Linguistics, Male, Middle Aged, Pattern Recognition, Automated, Sensitivity and Specificity, Speech, Speech Acoustics, Speech Production Measurement, Biomarkers, tau Proteins, Aphasia, Primary Progressive
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
