Health Disparities Arising from Artificial Intelligence Pilot Awards

• Data Bias: AI models trained on non-representative data may not perform effectively across different demographic groups, leading to misdiagnoses or ineffective treatments.
• Algorithmic Bias: AI algorithms can inherit biases present in their training data, resulting in unequal healthcare recommendations or eligibility in health research.
• Access Inequality: Disparities in access to AI technologies can widen the healthcare gap between different socioeconomic and racial groups.

The aim of this funding opportunity is to stimulate new research that investigates the ways in which AI contributes to health disparities and develops strategies to mitigate these effects. To be responsive to this opportunity, projects must include one or more of the following topics in the context of AI: identify contributors to disparities, develop mitigation strategies, enhance accessibility, and inform policy and practice. Because this funding opportunity is for pilot studies, applications should include an explicit discussion of how the intended aims can lead to larger studies in the future.

Examples of projects that would fit this opportunity include:

• Identify contributors to disparities:
  • Analyze how AI applications may contribute to health disparities.
  • Investigate the role of biased data and algorithms.
• Develop mitigation strategies:
  • Create methods to detect and correct biases in AI systems.
  • Propose frameworks for equitable AI development and deployment.
• Enhance accessibility:
  • Explore ways to improve access to AI-driven healthcare for underserved populations.
  • Assess the impact of AI technologies on different demographic groups.
• Inform policy and practice:
  • Provide evidence-based recommendations for policymakers and healthcare providers.
  • Promote ethical standards and best practices in AI healthcare applications.

CTSI Assistance (Optional)

If you would like to request a consultation, please email Aleks Zivic (abz17@pitt.edu). Make note of your intention to apply for the Health Disparities AI Pilot Awards in the subject line, and we are happy to provide feedback about the program or your application.

Award funding of up to $50,000 is available to cover direct costs; no indirect support will be provided. The award period will last for 12 months, beginning when all regulatory and administrative approvals have been received.

Projects must start within 3 months of the award notification.

Please be aware, under no circumstances, do the Health Disparity AI pilots have any mechanism for no-cost extensions; any funds that are not spent during the award period will be forfeited.

Before any funding can begin, awardees must provide documentation of all necessary regulatory approvals (IRB, IACUC, hSCRO, IBC, CORID, etc.). Once regulatory documentation is provided, awarded projects will undergo an administrative review from NCATS, which may take up to 30 days. Funding cannot begin until projects have been approved by NCATS. Because of this, all applicants are strongly encouraged to have the necessary regulatory documents ready for submission.

Applications that deliver a meaningful research experience to an early career investigator from a group underrepresented in the biomedical workforce may be eligible for an additional $5,000 of funding. Please see the full RFP For more details.