Northwestern University (Advanced AI in Medicine and Physics Lab)
Domain-Shuffle Temporal Attention Network for coronary vessel extraction from X-ray coronary angiography (XCA), trained entirely on synthetic temporal XCA data without requiring manual vessel annotations. By leveraging synthetic data generation and a domain-shuffle temporal attention mechanism, DOSTA-Net avoids the need for costly expert-labeled real angiography sequences while still learning temporally consistent vessel segmentation across frames of an XCA sequence.
Architecture
Hybrid
Domain-Shuffle Temporal Attention Network trained on synthetic temporal X-ray coronary angiography sequences (no manual annotation required) for coronary vessel extraction
Framework
PyTorch
Added to catalog
2026-08-14
Synthetic temporal X-ray coronary angiography dataset (DOSTA-Net)
Synthetically generated temporal XCA sequences used to train coronary vessel segmentation entirely without manual expert annotations.
Coronary vessel segmentation across frames of an X-ray coronary angiography sequence