VectorInstitute/odyssey
A toolkit for developing foundation models using Electronic Health Record (EHR) data.
This toolkit helps medical researchers and data scientists build specialized AI models using patient Electronic Health Records (EHR) data. It takes raw clinical data, like that from the MIMIC-IV dataset, processes it into a structured format, and then uses it to train powerful 'foundation models' that can predict patient outcomes or analyze health trends. The output is a trained AI model ready for specific clinical prediction tasks.
Use this if you are a healthcare data scientist or researcher working with large Electronic Health Record datasets and need to develop advanced AI models for clinical prediction or analysis.
Not ideal if you are looking for an out-of-the-box clinical prediction model or if you do not have access to large, raw EHR datasets and the technical expertise to manage them.
Stars
50
Forks
13
Language
Python
License
Apache-2.0
Category
Last pushed
Mar 10, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/transformers/VectorInstitute/odyssey"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related models
ycq091044/BIOT
BIOT - A framework for pretraining biosignals at scale. Large EEG pre-trained models.
AntixK/PyTorch-Model-Compare
Compare neural networks by their feature similarity
woodRock/fishy-business
Machine Learning for Rapid Evaporative Ionization Mass Spectrometry for Marine Biomass Analysis...
soda-inria/carte
Repository for CARTE: Context-Aware Representation of Table Entries
harryjdavies/HeartGPT
Interpretable Pre-Trained Transformers for Heart Time-Series Data