Storyteller in ADNI4: Application of an early Alzheimer's disease screening tool using brief, remote, and speech‐based testing
Bibliographic record
Abstract
INTRODUCTION: Speech-based testing shows promise for sensitive and scalable objective screening for Alzheimer's disease (AD), but research to date offers limited evidence of generalizability. METHODS: Data were taken from the AMYPRED (Amyloid Prediction in Early Stage Alzheimer's Disease from Acoustic and Linguistic Patterns of Speech) studies (N = 101, N = 46 mild cognitive impairment [MCI]) and Alzheimer's Disease Neuroimaging Initiative 4 (ADNI4) remote digital (N = 426, N = 58 self-reported MCI, mild AD or dementia) and in-clinic (N = 57, N = 13 MCI) cohorts, in which participants provided audio-recorded responses to automated remote story recall tasks in the Storyteller test battery. Text similarity, lexical, temporal, and acoustic speech feature sets were extracted. Models predicting early AD were developed in AMYPRED and tested out of sample in the demographically more diverse cohorts in ADNI4 (> 33% from historically underrepresented populations). RESULTS: Speech models generalized well to unseen data in ADNI4 remote and in-clinic cohorts. The best-performing models evaluated text-based metrics (text similarity, lexical features: area under the curve 0.71-0.84 across cohorts). DISCUSSION: Speech-based predictions of early AD from Storyteller generalize across diverse samples. HIGHLIGHTS: The Storyteller speech-based test is an objective digital prescreener for Alzheimer's Disease Neuroimaging Initiative 4 (ADNI4). Speech-based models predictive of Alzheimer's disease (AD) were developed in the AMYPRED (Amyloid Prediction in Early Stage Alzheimer's Disease from Acoustic and Linguistic Patterns of Speech) sample (N = 101). Models were tested out of sample in ADNI4 in-clinic (N = 57) and remote (N = 426) cohorts. Models showed good generalization out of sample. Models evaluating text matching and lexical features were most predictive of early AD.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.015 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".