Comprehensive Machine Learning-Enabled Outcome Prediction for Patients With Coronary Artery Disease Using Multicentre Patient Data
Bibliographic record
Abstract
BACKGROUND: Treatment decision-making for patients with coronary artery disease (CAD) can benefit from accurate patient outcome prediction. Although previous studies have used machine learning (ML) to develop prediction models, they were mostly on the basis of small patient cohorts with strict inclusion and exclusion criteria, limited features, and only internal validation. We aimed to develop and externally validate ML models to predict short- and long-term outcomes for patients with obstructive CAD using large-scale multicentre patient data. METHODS: We used a comprehensive data set from patients with obstructive CAD who underwent coronary angiography at 3 hospitals in Alberta, Canada between 2009 and 2019. To predict all-cause mortality and major adverse cardiovascular events at 90 days, 1 year, 3 years, and 5 years, > 12,000 features were considered in an extensive ML framework. In addition to traditional ML models, we used a generative transformer-based tabular foundation model, Tabular Prior-Data Fitted Networks (TabPFN; Prior Labs, Freiburg im Breisgau, Germany). To study real-time feasibility, secondary analyses limited feature sets to commonly available preangiography data. RESULTS: A total of 44,462 catheterizations from 38,767 patients were included. The median areas under the receiver operating characteristic curves of the best models, mostly TabPFNs, in external validation ranged from 0.796 to 0.845 and 0.694 to 0.755 for mortality and major adverse cardiovascular events, respectively. The minimum deployable preangiography feature set led to slightly lower but still reasonable performance. CONCLUSIONS: The large sample size, extensive feature set, external validation, and transformer architecture led to personalized models with robust prediction performance. Our models have the potential to improve CAD treatment decision-making via accurate prognosis.
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 | 0.002 | 0.008 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| 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".