Deep learning-based patient stratification for prognostic enrichment of clinical dementia trials
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Dementia probably due to Alzheimer's disease is a progressive condition that manifests in cognitive decline and impairs patients' daily life. Affected patients show great heterogeneity in their symptomatic progression, which hampers the identification of efficacious treatments in clinical trials. Using artificial intelligence approaches to enable clinical enrichment trials serves a promising avenue to identify treatments. In this work, we used a deep learning method to cluster the multivariate disease trajectories of 283 early dementia patients along cognitive and functional scores. Two distinct subgroups were identified that separated patients into 'slow' and 'fast' progressing individuals. These subgroups were externally validated and independently replicated in a dementia cohort comprising 2779 patients. We trained a machine learning model to predict the progression subgroup of a patient from cross-sectional data at their time of dementia diagnosis. The classifier achieved a prediction performance of 0.70 ± 0.01 area under the receiver operating characteristic curve in external validation. By emulating a hypothetical clinical trial conducting patient enrichment using the proposed classifier, we estimate its potential to decrease the required sample size. Furthermore, we balance the achieved enrichment of the trial cohort against the accompanied demand for increased patient screening. Our results show that enrichment trials targeting cognitive outcomes offer improved chances of trial success and are more than 13% cheaper compared with conventional clinical trials. The resources saved could be redirected to accelerate drug development and expand the search for remedies for cognitive impairment.
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.
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 | 0.005 | 0.007 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| 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) | 0.000 | 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 it