.png&w=3840&q=75)
.png&w=3840&q=75)
Patients admitted to the emergency department and the intensive care unit.
Datathon · 4th Annual
MD+ 2025 Datathon Empowering Patients Through AI.
A national, month-long event bringing medical trainees together to explore complex healthcare challenges and build data-driven solutions on real clinical datasets.
- Dates
- 10/17/25 – 11/19/25
- Location
- Remote
- Team size
- 3–6
- Open to
- Med students · Residents · Grad students
About
The 4th Annual Datathon.
The 4th annual MD+ Datathon was a national, month-long event that brought medical students from various backgrounds together to explore complex healthcare challenges and create data-driven solutions. Hosted by MD+ and its partners, the datathon encouraged interdisciplinary collaboration, innovative thinking, and real-world problem solving using clinical datasets. The theme was Empowering Patients Through AI.
From the datasets provided, participants were asked to identify an addressable problem to explore by analyzing the dataset, then create an actionable recommendation or product.
Why “Empowering Patients Through AI”?
Students were encouraged to interpret the theme broadly and creatively — health literacy, access to care, after-visit summaries, chatbots for chronic conditions, insurance navigation, language accessibility, and more.
Who should join?
Medical students, residents, and graduate students pursuing degrees related to human health. Teams of 3–6 trainees; multidisciplinary teams are highly encouraged.
What you'll gain
Skills in data analysis and machine learning, experience tackling high-impact problems in patient care, opportunities to publish — and monetary prizes for the winners.
Datasets
What teams worked with.
Physiological, hormonal, and self-reported events and symptoms for menstrual health tracking with wearables.
Usage requirements
Daily patient tracking of symptom severity, treatments and doses, and potential environmental triggers (foods, stress, allergens).
Beginner track · Direct CSV download via Kaggle
441,455 individuals and 330 features across survey questions and calculated variables, with three diabetes classes (none/pregnancy-only, prediabetes, diabetes).
Beginner track · Direct CSV download via Kaggle
Timeline
Key dates.
- Anytime
- Fri 10/17/25 · 11:59pm AOE
- Mon 10/20/25 · 7pm EST
Kickoff meeting
- Mon 10/20/25 · 11:59pm AOE
Team registration deadline
Register your team. If you haven't found teammates, reach out to us!
- Wed 10/29/25
Week 2 check-in
Our team reaches out via Slack for check-ins.
- Wed 11/5/25
Week 3 check-in
- Thu 11/13/25 · 11:59pm AOE
Final submission
- Wed 11/19/25 · 7pm EST
Pitch competition
Partners + sponsors
Backed by serious operators.
Gold sponsors
- Doximity
- Inflo Health
Bronze sponsors
- Offcall
Research collaborators
- AMSA
- ConductScience
FAQs
Common questions.
I have little/no data science or computational experience. Can I still participate?
Absolutely! Learning the computational tools is half the fun. Participants have access to tutorials covering the basics of Python and R and how to analyze the dataset. Submissions are judged on more than technical complexity — judges care more about the insights derived and the data analysis than the computational novelty.
What is the time commitment?
We encourage all participants to attend the kickoff and the pitch competition, but otherwise there are no required synchronous events. The time commitment is as much as you and your team want to put into the project.
What is the final product we are submitting?
Your team submits a report (of any length) of your findings along with your code. The report should include Background, Methods, Results, Discussion, and Conclusions.
Do I need a team beforehand?
No. Search for teammates by adding your name and bio to the shared slides.
Can I see previous examples of projects?
Yes! In 2024 the theme was “Responsible Generative AI,” and teams could publish an abstract-version of their findings in PRISM's special-edition journal. In 2023 the theme was Value-Based Care.
Questions? Reach the organizers
Other editions