Predicting the future risk and outcomes of severe heart failure and coronary artery disease with machine learning in the UK Biobank Cohort
Bibliographic record
Abstract
BACKGROUND: In order to seriously impact the global burden of heart failure (HF) and coronary artery disease (CAD), identifying at-risk individuals as early as possible is vital. Risk calculator tools in wide clinical use today are informed by traditional statistical methods that have historically yielded only modest prediction accuracy. METHODS: This study uses machine learning algorithms to generate predictions models for the development and progression of severe HF and CAD. Participants (~485,000 followed in the UK Biobank over 7 years) were stratified by cardiac status at the time of enrollment (asymptomatic, high-risk and affected); separate prediction models were built for each stratum. Participants were split between a training set (80%) and holdout dataset (20%), all performance metrics are reported for the holdout dataset. RESULTS: Out of 6 machine learning algorithms screened, artificial neural networks (ANN) most successfully predicted future disease across the various strata (area under the curve: 0.77-0.86 for 10/12 models), results were very consistent between methodologies. Models trained using ANN showed excellent calibration in all strata and across the entire spectrum of risk (0.4-1.2% average observed/predicted difference across 10 deciles of risk). Key predictive features included age, frailty, adiposity, history of hypertension and diabetes, tobacco use and family history of heart disease and were consistent between models for HF and CAD. CONCLUSIONS: When deployed as a patient-facing application, the prediction models presented here will be able to provide both user-specific predictions and simulate the effect of changes in lifestyle and of prophylaxis interventions, thus resulting in an individualized patient counselling and management tool.
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 | 0.005 | 0.016 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 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) | 0.002 | 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".