Submit a Model

Add a cardiovascular AI model to the catalog. All submissions are reviewed before they appear publicly.

gavel What we accept

Pre-trained deep learning models processing ECG, echocardiography, cardiac MRI/CT, cardiology text, or multimodal cardiovascular data. See our detailed acceptance criteria here.
Public training and inference code.
Downloadable model weights, or a justification for why they cannot be made public.
Not raw patient datasets, live clinical APIs, or general (non-cardiovascular) medical AI.
A single model (only one set of weights). If you have multiple versions of the same model, please only submit the one you think would be most useful to other researchers.
An accompanying paper (or preprint) describing the model's function.

I confirm this model meets the acceptance criteria above.

description Basic info

Model name * Version Creator / organization *

Description *

code Sources

Code source URL * Model availability * Weights are publicly available Weights are not public

Model files are hosted at the same URL as the code

Model files URL *

help

Please link to the repository or location your model weights are stored, such as a GitHub or Hugging Face repo, Zenodo page, or Google Drive folder. If your model is stored in a place with other models, link to the directory/page with only your model. If it is stored somewhere that only includes your model (multiple versions are fine), you can just link to there.

License *

Private weight justification *

database Training & evaluation data

Select the dataset(s) used to train or evaluate this model, and what each was used for. Choose "Other" if a dataset isn't listed, it'll be flagged for the catalog team to review.

Additional notes optional

description Publication

Paper title * Paper URL Year DOI

science Outputs & performance

Add one entry per task the model performs (e.g. "AFib detection"), along with any performance metrics you can report for it.

public Your contact info optional

Your name Your email

CVAI Catalog

View Catalog