Machine Learning Approach for Predicting Amyloid and Tau Positivity in Alzheimer's Disease Using Clinically Accessible Features
Bibliographic record
Abstract
Abstract Background Prediction of Alzheimer's disease (AD) biomarkers can improve public health strategies, especially if achieved with easily collectable data in a single consultation. Machine learning (ML) offers versatile tools for clinical and research applications. This study investigated a ML model's ability to predict amyloid and tau positivity using easily obtainable features. Method Individuals with amyloid and tau status were selected from ADNI, TRIAD, and PPMI datasets (Model Building) and from the NACC dataset (Validation). Shared clinical features included age, sex, education, clinical diagnosis, MoCA scores, and BMI. Amyloid positivity was defined by amyloid‐PET (PIB‐PET, FBB‐PET, or AZD4694‐PET) or CSF AB42, and tau positivity by Tau‐PET (MK6240‐PET, AV1451‐PET) or CSF p ‐tau181. For NACC, positivity derived from fields AMYLPET and TAUPETAD. Data processing is summarized in Figure 1. Result The Model Building sample included 1593 individuals (mean MoCA 25.3 ± 4.3). The Validation cohort included 861 individuals (mean MoCA 22.2 ± 2.8). The model achieved high performance for predicting amyloid and tau positivity, with mean AUCs of 0.89 and 0.83 for Model Building and Validation, respectively (Figure 2a, 2b). In Validation, high sensitivity (0.93) came at the expense of specificity (0.65), while Model Building showed balanced sensitivity and specificity (0.81 and 0.83). Higher age, lower MoCA, cognitive impairment, female sex, and lower BMI increased the probability of positivity (Figure 2c, 2d). Cognitive impairment was the most impactful feature in both Model Building and Validation datasets, followed by MoCA and age (Figure 2e, 2f). Conclusion Predicting AD biomarkers using ML and readily collectable features is feasible and accurate. The model's high sensitivity indicates a potential research utility in clinical trials for population screening to minimize false negatives. Future efforts should enhance generalizability, explore additional features, and prioritize real‐world validation.
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 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| 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) | 0.001 | 0.001 |
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".