PyHealth Research Initiative

Overview

The PyHealth Research Initiative is a year-round, open research program that brings together talented individuals from diverse backgrounds—including students, engineers, researchers, and healthcare professionals—to conduct cutting-edge research in healthcare artificial intelligence. This inclusive program welcomes anyone passionate about advancing computational healthcare, regardless of their career stage or institutional affiliation. Through this program, participants work on innovative projects that advance the field of computational healthcare, contributing to publications, open-source software, and the broader healthcare AI community.

Our goals are to build:

1. Easily accessible and reproducible research — Making healthcare AI research transparent and replicable

2. Solutions to real-world healthcare problems — Tackling important clinical challenges with practical impact

3. Connections with healthcare professionals — Bridging the gap between AI researchers and clinical practitioners

The initiative provides participants with hands-on experience in:

• Healthcare AI Research: Working on real-world healthcare problems using electronic health records (EHRs) and clinical data

• Machine Learning Development: Implementing and evaluating state-of-the-art deep learning models

• Open-Source Contribution: Contributing to the PyHealth library and ecosystem

• Academic Publishing: Co-authoring research papers and presenting findings

• Collaborative Research: Working alongside researchers and industry partners

Program Logistics

Format: Remote

Time Commitment: 10–20 hours per week during active research cycles

Eligibility: Open to anyone! We welcome people from all backgrounds—students, engineers, researchers, and healthcare professionals. We don’t care about your title or institution. We only ask that you have the ability to write decent-quality code and are self-driven to work hard on healthcare problems. See How to Apply to get started.

Open Projects: Browse available research projects and find one that matches your interests: Open Projects List

The program runs on a rolling, year-round basis with recurring terms aligned to major healthcare AI conference cycles. Each term culminates in a submission to a top-tier venue.

Upcoming Research Terms

Term	Period	Target Conference	Est. Submission Deadline
Summer Term	Apr – Aug 2026	ML4H 2026	~Sep 2026
Fall Term	Sep – Dec 2026	CHIL 2027	~Feb 2027
Spring Term	Jan – Apr 2027	MLHC 2027	~May 2027

Research Contributions

Below is a comprehensive list of research contributions from PyHealth Research Initiative participants over the years. These projects span various topics including clinical prediction, model interpretability, drug recommendation, and healthcare AI infrastructure.

PyHealth Research Initiative Projects

Year	Researcher(s)	Paper Title	Venue	Links
2026	Arjun Chatterjee, Sayeed Sajjad Razin	Making Conformal Predictors Robust in Healthcare Settings: a Case Study on EEG Classification	Under Review at AIME	Paper
2025	Zilal Eiz Al Din	MIMIC-RD: Can LLMs differentially diagnose rare diseases in real-world clinical settings?	ML4H 2025	Paper
2025	Abraham Francisco Arellano Tavara, Umesh Kumar	Prostate-VarBench: A Benchmark with Interpretable TabNet Framework for Prostate Cancer Variant Classification	ML4H 2025	Paper
2025	Sharim Khan	Social Determinants of Health Prediction for ICD-9 Code with Reasoning Models	ML4H 2025	Paper

Note

This table is continuously updated as new research is published.

Latest from ML4H 2025: Our first cohort successfully published three papers at ML4H, covering rare diseases, social determinants of health, and prostate cancer genomics—all at the forefront of healthcare and AI research!

Research Areas

PyHealth Research Initiative projects span a wide range of healthcare AI domains, including but not limited to:

Healthcare Data Modalities

• Electronic health records (EHRs) and clinical notes

• Medical imaging (X-rays, MRI, CT scans)

• Physiological signals (EEG, ECG, biosensors)

• Genomic and molecular data

• Multi-modal healthcare data integration

Model Development

• Novel deep learning architectures for healthcare

• Foundation models and transfer learning

• Interpretable and explainable AI methods

• Uncertainty quantification and calibration

• Survival analysis and time-to-event modeling

Real-World Applications

• Clinical decision support systems

• Drug discovery and repurposing

• Patient risk stratification

• Healthcare resource optimization

• Personalized treatment recommendations

How to Apply

Interested in participating in the PyHealth Research Initiative? Here’s the three-step application process:

Step 1: Join the PyHealth Discord

Join our Discord server — This is where all official announcements and communication happen. It’s the best way to connect with the team, ask questions, and stay updated.

Step 2: Submit an Application Form

Especially important for those interested in industry collaborations: Application Form

Step 3: Submit a High-Quality Pull Request to PyHealth

Applications are evaluated based on your contributions to PyHealth. Submit a high-quality pull request (PR) to the PyHealth repository. Strong PRs that demonstrate understanding of the codebase, good software engineering practices, and meaningful contributions will be considered for the program.

Need ideas for contributions?

• Bounty List: Check our curated list of contribution opportunities: Bounty List

• Contribution Guide: Learn how to contribute: How to Contribute

• GitHub Issues: Browse open issues in the repository for contribution ideas

What You’ll Gain

By participating in the PyHealth Research Initiative, you’ll gain:

1. Publication: A peer-reviewed paper published at a respectable venue, establishing you as a credible researcher in the field

2. Hands-on Experience: Direct experience working on real-world healthcare AI problems

3. Open-Source Impact: Code merged into PyHealth that can be leveraged for other career goals

4. Mentorship: Direct mentorship from experts across a wide variety of healthcare disciplines

5. Collaborations: Potential industry and academic collaborations

6. Challenging Problems: Work on interesting and impactful healthcare challenges

Past Participants

The PyHealth Research Initiative has brought together participants from diverse backgrounds and institutions around the world, including:

• University of Illinois Urbana-Champaign

• Georgia Tech

• Industry professionals and independent researchers

• And many more…

Participants and alumni of the program have gone on to pursue graduate studies, research positions, and careers in healthcare AI at leading institutions and companies.

Stay Connected

For more information about the PyHealth Research Initiative:

Best way to reach us: Join our Discord server — Join the community, ask questions, share what you’re working on, and get the fastest response from the team!

Additional resources:

• GitHub: https://github.com/sunlabuiuc/PyHealth — Star the repo, open issues, submit PRs

• Documentation: https://pyhealth.readthedocs.io

• Mailing List: Subscribe here — Get updates delivered to your inbox

• Email: johnwu3@illinois.edu

Acknowledgments

The PyHealth Research Initiative is made possible through the support of:

• SunLab at the University of Illinois Urbana-Champaign

• National Science Foundation (NSF)

• Industry partners and collaborators

We are grateful to all mentors, participants, and collaborators who have contributed to the success of this program.