Prediction of incident cardiovascular events using machine learning and CMR radiomics
Problem
Predicting whether patients will develop atrial fibrillation, heart failure, myocardial infarction, or stroke before symptoms occur.
Why it matters
Early prediction of major cardiovascular events could enable preventive treatment and save lives.
What the AI did
Built seven integrative machine learning models per disease outcome using cardiovascular magnetic resonance radiomics features, CMR indices, and vascular risk factors from UK Biobank participants, classified via support vector machine.
Results
not stated in source (abstract truncated before performance results)
Impact
Feasibility of radiomics-plus-ML for incident CVD prediction was evaluated; no performance or clinical adoption figures stated.
Record
| Venue | European Radiology |
|---|---|
| Published | 2022 |
| Work type | article |
| Authors | Esmeralda Ruiz Pujadas, Zahra Raisi‐Estabragh, Liliána Erzsébet Szabó, Celeste McCracken, Cristian Izquierdo, Víctor M. Campello, Carlos Martín-Isla, Angélica Atehortúa et al. |
| Citations | 47 (OpenAlex, at import time) |
| References | 43 |
| Model or method | Support vector machine |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1007/s00330-022-09323-z |
| Open access | hybrid |
| Keywords | machine learning, CMR radiomics, cardiovascular disease, UK Biobank, support vector machine, risk prediction |
Importance rating
significant — thousands to millions affected
Early prediction of AF, heart failure, MI, and stroke could enable prevention; evaluated on UK Biobank participants, but abstract gives no mortality/prevalence figures to justify higher score.