Introduction
Please review the hackathon rules and code of conduct before the event (see the Rules tab on Devpost).
The detailed judging criteria is available here.
We encourage competitors to seek advice and guidance from the mentors and organizers who will be available throughout the event as you develop your solution (more details coming soon).
Tools and Technologies
A list of open/public datasets is available here. You may also use publicly available datasets not on this list for the competition!
Inspiration
Imagine you oversee surveillance or epidemiology at a government organization, such as the Centers for Disease Control and Prevention (CDC) or a state/county/local health department. There’s currently an STI epidemic and you’ve been limited to your current solutions but need new approaches to stop the epidemic. What should you do? How can you find a new surveillance method or scalable technology-based intervention tool to address the epidemic?
Whether you want to work with raw data and build unique models or take a new approach to address limitations in STI surveillance (link) and improve existing approaches, the sky is the limit. For example, you might seek to find new relationships between substance use and STI cases, or STI cases among understudied populations. Judges will evaluate the novelty of your approach, too, so you might consider novel data sources (link) not typically used in the STI sphere (social media data; internet and mobile cell phone use to identify data about social behavior, including the types of searches performed on the Web; information communicated to schools and social networks; the types and sources of medicines bought by individuals, etc.). This information could lead to early warning or alert systems in health care that would enable public health sectors to identify and track disease outbreaks or meet other types of needs (link).
STI interventions are limited by the humans who deliver them. Social/behavior interventions (link) such as motivational interviewing, contingency management, and community peer leader interventions can increase STI prevention, but they require a lot of time, financial resources, and often rely on more human involvement than is feasible (especially in jurisdictions with limited resources, such as small, rural counties or clinics). Advances in AI (link) have made it possible to scale human work in ways not previously possible. AI can find patterns that would have taken humans years to find, or help to rapidly interview or recommend treatments to a large group of patients that were not possible before due to logistical and scalability issues.
Embrace the capabilities of AI to address complex challenges around STI prevention and control. We are encouraging broad learning and experimentation with AI powered software tools like Large Language Models (ChatGPT) and other creation tools like Dall-E, Midjourney and others, but the toolset is up to you. Examples of how AI approaches could be used to address STIs include: disease diagnosis (including medical image analysis); personalized medicine (for example, using AI to implement a culturally and appropriate sexual history taking that could be integrated into a patient's electronic health record); physician decision support software; assisted self-interviewing; patient monitoring; connecting patients to trusted care sources; providing access to age appropriate, accurate, and trusted STI information/resources; or improve STI surveillance (AI might be used to analyze patterns among electronic health records and insurance claims data to assess risk factors for STIs and assess how they might vary among demographic groups to tailor intervention efforts). Another approach is immersive virtual reality applications (link) might use AI to increase understanding of experiences and the stigma of people who have contracted STIs. AI-informed virtual reality (ie, AI to learn from the experiences of each virtual reality user) has the potential to create an engaging and empathy-building educational environment.
As AI is still unexplored, especially in areas with patient sensitivities such as STI prevention and care, it is essential that teams are able to describe the potential ethical and implementation issues with these approaches and a realistic assessment of how they would overcome them. For example, for various reasons, many patients might petition against the use of AI to provide them medical recommendations instead of being able to see their doctor; many providers might object to being given recommendations from AI. Teams must have gained a pulse on public opinion on this topic to develop an implementable solution, and/or focus on a population or region that is more accepting of novel solutions such as AI/technology.
Participants should develop a functional prototype or a proof of concept that demonstrates their solution. The submission will include both a pitch presentation (e.g., video) and a link to the technology or model that you/your team developed (see “What to Submit” section below for more details).
For example, if you were to demonstrate the role of social media data in improving STI surveillance, you would need to be able to describe the model, how and why you think it works, demonstrate the improvement it has in performance compared to the current data analytics/surveillance approach (if there is one), and show a visualization of your results (for example, over time, if it is a longitudinal analysis) (see “What to Submit” section below for more details).
For information on STIs, review the following resources:
- Primer on STIs and data needs is available here.
- The recent National Academies report: Sexually Transmitted Infections: Adopting a Sexual Health Paradigm, here.
- CDC has a wealth of information on STIs, here.
Frequently asked questions
Does the STI Hackathon cost money to participate in?
No! Thanks to the generosity of our sponsors, the event is free for all participants.
What if I have other questions?
To submit a question, please send an email to [email protected] with the subject line “Question – STI Hackathon.” Staff will respond to you within 3 business days if your question has not been answered on this posting.
What will the grants entail for those who are seletected?
Grantees will be asked to continue working to improve their prototype for at least 4 months and hold a 30-minute check-in call with National Academies staff (and potentially other mentors) once a month within that period to discuss progress and barriers. At the end of the 4 month period, grantees will submit a 2-page document summarizing what has been accomplished to date, barriers that still need to be overcome, and next steps—if any.
Contact the organizers at [email protected]
As we receive questions from you, we will post them to a discussion thread in the discussion forum so all competitors benefit from the answers.