Speech timing metrics
1 paper4 findings
- Humans1
Associations
4Alzheimer dementia
upin silico1
1 study
Alzheimer dementia
upin silico1
Speech timing metrics predict alzheimer dementia in silico (AUC = .75).
“machine learning models trained on English‐speaking AD patients and HCs generalized best onto Spanish speakers when fed with speech timing than with word‐level features, both for identifying patients (AUC = .75)”
In silicocross-linguistic machine learning model trained on English speakers tested on Spanish speakers
AAIC 2025 abstract7 Jan 2026
Self-rated reading abilities
downin humans1
1 study
Self-rated reading abilities
downin humans1
Speech timing metrics are associated with self-rated reading abilities in humans (r = -.52, p = .03).
“as well as self‐rated reading abilities in Portuguese‐speaking AD patients ( r = ‐.52, p = .03).”
Humansin Portuguese-speaking AD patients
AAIC 2025 abstract7 Jan 2026
ACE-III
upin humans1
1 study
ACE-III
upin humans1
Speech timing metrics predict ACE-III in humans (r = .63, p < .001).
“and Spanish‐speaking (ACE‐III: r = .63, p < .001) cohorts, as well as self‐rated reading abilities in Portuguese‐speaking AD patients”
Humansin Spanish-speaking cohort (early-stage AD and HCs)
AAIC 2025 abstract7 Jan 2026
CERAD-DR
upin humans1
1 study
CERAD-DR
upin humans1
Speech timing metrics predict CERAD-DR in humans (r = .73, p < .001).
“Speech timing metrics predicted cognitive status performance in HCs and AD patients from Kiswahili‐speaking (CERAD‐DR: r = .73, p < .001)”
Humansin Kiswahili-speaking cohort (early-stage AD and HCs)
AAIC 2025 abstract7 Jan 2026