A novel deep network with triangular-star spatial-spectral fusion encoding and entropy-aware double decoding for coronary artery segmentation
Framework
The framework of this work offers a visual overview that delineates the entire flow from problem statement to methodology, and experimental setup.
Quantitative Comparison
Quantitative comparison with state-of-the-art methods on the CTA119 dataset
and the ASOCA dataset The best results are in bold, and the second-best results are
underlined.
Qualitative Comparison
Qualitative comparison of three typical cases between different methods for
coronary artery segmentation. The yellow and green dashed circles highlight the regions
for better visual comparison.
Usage
Data preparation
ASOCA dataset:
https://asoca.grand-challenge.org
TubeTK dataset:
https://public.kitware.com/Wiki/TubeTK/Data
Your datasets directory tree should be look like this:
data
├── npy
├── img
├── 1.npy
├── 2.npy
└── ...
└── mask
├── 1.npy
├── 2.npy
└── ...
Training
python train.py
Testing
python test.py