Amsterdam UMC (Zhai et al.)
Model weights not public. Contact creators for more information.
Coronary artery calcium (CAC) scoring model that transfers a CNN trained for calcium scoring on non-contrast CT (NCCT) to coronary CT angiography (CCTA), where iodinated contrast otherwise confounds calcium detection and large annotated CCTA training sets are scarce. The CAC-scoring CNN is split into a feature generator and a classifier; the feature generator is trained on the NCCT source domain and adapted to the CCTA target domain via adversarial learning combined with a maximum-mean-discrepancy loss, while the source-domain classifier is reused unchanged for the target domain. Builds directly on the authors' earlier non-contrast CT calcium-scoring network.
Architecture
Hybrid
CAC-scoring CNN divided into a feature generator and a classifier; the feature generator is trained on NCCT (source domain) and adapted to CCTA (target domain) via adversarial learning and a maximum-mean-discrepancy loss, reusing the source-domain classifier unchanged
Framework
PyTorch
Added to catalog
2026-08-12
Amsterdam UMC Mixed Non-Contrast CT / Coronary CTA Cohort (Zhai domain adaptation)
Paired non-contrast CT (source domain) and coronary CT angiography (target domain) scans used to train and adapt a coronary artery calcium scoring CNN via adversarial domain adaptation, building on the authors' earlier non-contrast CT calcium-scoring cohort.
Coronary artery calcium detection/scoring on coronary CT angiography (CCTA), using a feature generator adapted from non-contrast CT via unsupervised domain adaptation