Predicting conversion in cognitively normal and mild cognitive impairment individuals with machine learning: Is the CSF status still relevant?
Bibliographic record
Abstract
INTRODUCTION: Machine learning (ML) helps diagnose the mild cognitive impairment-Alzheimer's disease (MCI-AD) spectrum. However, ML is fed with data unavailable in standard clinical practice. Thus, we tested a novel multi-step ML approach to predict cognitive worsening. METHODS: We selected cognitively normal and MCI participants from the Alzheimer's Disease Neuroimaging Initiative dataset and categorized them on total tau/amyloid beta 1-42 ratios. ML was applied to predict the 3-year conversion with standard clinical data (SCD), assess the model's accuracy, and identify the role of cerebrospinal fluid (CSF) biomarkers in this approach. Shapley Additive Explanations (SHAP) analysis was carried out to explore the automated decisional process. RESULTS: The model achieved 84% accuracy across the entire cohort, 86% in patients with negative CSF, and 88% in individuals with AD-like CSF. SHAP analysis identified differences between CSF-positive and -negative patients in predictors of conversion and cut-offs. CONCLUSIONS: The approach yielded good prediction accuracy using SCD. However, CSF-based categorizations are needed to improve predictive accuracy. HIGHLIGHTS: Machine learning algorithms can predict cognitive decline with standard and routinely used clinical data. Classification according to cerebrospinal fluid biomarkers enhances prediction accuracy. Different cut-offs could be applied to neuropsychological tests to predict conversion.
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 | 0.008 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 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 | 0.000 |
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".