Atrial fibrillation detection leaderboard
Benchmark results for atrial fibrillation detection from PPG and wearable sensors.
AFib / arrhythmia
Dataset: afppgecg · Metric: Bal. accuracy ↑
| Rank | Method | Bal. accuracy |
|---|---|---|
| #1 | VitalAgent | 0.672 |
AFib / arrhythmia
Dataset: external-single-lead-af · Metric: AUROC ↑
| Rank | Method | AUROC |
|---|---|---|
| #1 | ECGFounder | 0.957 |
AFib / arrhythmia
Dataset: icentia11k · Metric: Bal. accuracy ↑
| Rank | Method | Bal. accuracy |
|---|---|---|
| #1 | VitalAgent | 0.678 |
AFib / arrhythmia
Dataset: liu-ppg-arrhythmia-6class · Metric: AUROC ↑
| Rank | Method | AUROC |
|---|---|---|
| #1 | Pulse2View | 0.9697 |
AFib / arrhythmia
Dataset: mimic-iii-ext-ppg · Metric: Macro-F1 ↑
| Rank | Method | Macro-F1 |
|---|---|---|
| #1 | PhysioSkeptic | 0.886 |
AFib / arrhythmia
Dataset: mimic-perform · Metric: Accuracy ↑
| Rank | Method | Accuracy |
|---|---|---|
| #1 | TF-MambaNet | 0.83 |
AFib / arrhythmia
Dataset: ptb-xl-rhythm · Metric: AUROC ↑
| Rank | Method | AUROC |
|---|---|---|
| #1 | MS-ECG-FM | 0.987 |
AFib / arrhythmia
Dataset: pulsewatch · Metric: AUROC ↑
| Rank | Method | AUROC |
|---|---|---|
| #1 | Development of an efficient deep learning model | 0.9703 |
AFib / arrhythmia
Dataset: pulsewatch · Metric: Macro-F1 ↑
| Rank | Method | Macro-F1 |
|---|---|---|
| #1 | Random Forest, engineered features | 0.842 |
AFib / arrhythmia
Dataset: stanford-ppg · Metric: AUROC ↑
AFib / arrhythmia
Dataset: stanford-ppg · Metric: Macro-F1 ↑
| Rank | Method | Macro-F1 |
|---|---|---|
| #1 | ECG-PPG JEPA 345M | 0.834 |