Doppler-angle estimation on Rotation-augmented image corpus Patient-level sampling
5.93MAE (°)5-model ensemble (stacked)
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| 5-model ensemble (stacked)Hyperparameter Tuning=Optuna-tuned, Aggregation=stacked, Validation=pooled out-of-fold2025.08 | 5.93 | 8.53 | 0.952 | |
| 5-model ensemble (mean)Hyperparameter Tuning=Optuna-tuned, Aggregation=pooled out-of-fold means2025.08 | 6.89 | 9.89 | 0.932 | |
| DenseNet201Hyperparameter Tuning=Optuna-tuned, Validation=5-fold cross-validation2025.08 | 8.62 | 10.8 | 0.886 | |
| ResNet50Hyperparameter Tuning=Optuna-tuned, Validation=5-fold cross-validation2025.08 | 9.79 | 13.3 | 0.842 | |
| VGG19Hyperparameter Tuning=Optuna-tuned, Validation=5-fold cross-validation2025.08 | 10.21 | 14.97 | 0.871 | |
| InceptionV3Hyperparameter Tuning=Optuna-tuned, Validation=5-fold cross-validation2025.08 | 10.66 | 14.66 | 0.856 | |
| XceptionHyperparameter Tuning=Optuna-tuned, Validation=5-fold cross-validation2025.08 | 10.88 | 14.76 | 0.851 |