IC‐P‐131: Robust Identification of Amnestic MCI Progressors to Probable Alzheimer's disease Via Baseline MRI Analysis
Bibliographic record
Abstract
MR images taken from different centers appear dissimilar due to a variety of scanner-dependent effects. This situation is particularly acute in large, multi-centric settings such as ADNI. Our goal was to assess the accuracy of prediction to progression to AD for ADNI MCI subjects for an automated classification technique (Duchesne et al, Neurobiol. Aging 2008). A total of 481 subjects were available for final analysis. The Volume of Interest (VOI) Group consisted in 75 probable AD and 75 age-matched controls from the LENITEM dataset. The Study Group consisted in 331 MCI subjects from ADNI (129 progressors (1.50 years (SD: 0.69 years)) and 202 stable). MRI data for the VOI Group were acquired in Italy on a 1.0T scanner; ADNI data were acquired on 56 different 1.5T scanners. To increase technique robustness, we added noise removal and intensity standardization to the previous image pipeline. Model features were local volume change and standardized intensity sampled in a pathology-specific VOI, defined as areas of grey matter differences between AD/controls in the VOI group (Figure 1). We randomly split the Study Group in a Model Group of 166 and a Test Group of 165 subjects. We generated a linear model of 134 normally distributed image features explaining 95% of data variance from the Model Group. We projected Test Group data in the model space, and assessed classification accuracy with forward, stepwise linear discriminant analysis (p-to-enter = 0.15) in a k-fold fashion (k = 10), averaged over 5 trials. We obtained 82.3% accuracy, 77.5% specificity, and 85.9% sensitivity with a median number of 37 variables in the discrimination function. We performed comparison studies using other publicly available data (ADAS-COG; hippocampal volumes; SPARE-AD) (these were not available for all subjects). Of note, classification based on SPARE-AD reached 71.7% accuracy on a subset of 61 MCI (Figure 2).
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| 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 teacher head, 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".