DeepCAC

Harvard AIM Lab (Zeleznik et al.)

Non-contrast cardiac CT

Filter catalog by Modality:
Cardiac CT

Coronary artery calcium (CAC) scoring

Filter catalog by Disease / Trait:
Coronary & Ischemic Disease

Segmentation

Filter catalog by Task Type:
Segmentation & Detection

CNN (3D)

Filter catalog by Architecture:
Convolutional (CNN)

GPL 3.0

Filter catalog by License:
Copyleft

code View code & weights

Fully automatic pipeline that localizes the heart, segments coronary calcium, and produces an Agatston-style coronary artery calcium score from gated and non-gated chest/cardiac CT. A three-stage 3D CNN performs each step in sequence. Validated across the Framingham, NLST, PROMISE, and ROMICAT-II cohorts, where the resulting calcium score predicted cardiovascular events with hazard ratios up to 4.3. Developed by the Harvard AIM Lab.

memory Specifications

gavel License

GPL 3.0

check_small Open source check_small Commercial use OK check_small Modifications allowed Attribution required Share-alike required

License for model weights only. Associated code may be licensed seperately, check code source for specific terms.

description Publication

Deep convolutional neural networks to predict cardiovascular risk from computed tomography open_in_new

2021 · original paper

DOI: 10.1038/s41467-021-20966-2

database Training & evaluation data

Framingham Heart Study CT1 Cohort (DeepCAC development)

train

public 1,636 subjects · USA · 2002-2005

Framingham Heart Study CT2 Cohort (DeepCAC validation)

validation

public 663 subjects · USA · 2008-2011

National Lung Screening Trial (NLST) Low-Dose Chest CT Cohort (Lessmann) open_in_new

train

public 1,744 subjects · USA

Low-dose chest CT scans from heavy smokers in the NLST lung-cancer screening trial, reconstructed with both soft and medium/sharp filters, with per-scan coronary artery, thoracic aorta, and cardiac-valve calcification annotations.

science Capabilities & performance