Developing AI Applications for Traumatic Brain Injury Care (2026-2027)

Background

Each year, an estimated 5 million Americans seek emergency medical care for traumatic brain injury (TBI), making it a major cause of disability and death, particularly among young adults. Long-term consequences of TBI can include mood and attention disorders, cognitive impairment and increased risk of suicide. Despite the seriousness of these outcomes, providers often struggle to determine an accurate prognosis. One survey found that only 37% of healthcare providers believed they could reliably predict outcomes for TBI patients, and two-thirds believed a better predictive model would improve their clinical decision-making.

Recent advances in machine learning have demonstrated the potential for artificial intelligence (AI) systems to improve clinical care by synthesizing large, complex healthcare data. While AI tools have begun to influence decision-making in other areas of medicine, TBI care remains particularly challenging due to the complexity of volumetric imaging, electronic health record data and diverse care pathways. For AI systems to be effective and responsibly implemented, developers must work closely with clinicians to understand workflows, identify meaningful use cases and ensure that tools align with the needs and expectations of frontline providers.

Project Description

This project team will build on an ongoing multi-year effort to develop, refine and evaluate an AI-driven clinical decision support (CDS) tool for TBI care. Building on prior work that identified potential applications for AI and produced a data-cleaning pipeline, the 2026-2027 team will focus on linking stakeholder perspectives with the creation and evaluation of a prototype CDS application. Collaborators are currently finalizing a model that will underlie the CDS tool, and students on this project will collaborate with technical stakeholders that are developing the model, clinical stakeholders involved in TBI care and collaborators helping to develop a prototype user interface for the CDS tool.

The team will be organized into two interconnected subgroups:

1. Qualitative research:

Team members will expand on prior research and conduct additional qualitative interviews with providers, including emergency physicians, neurologists, neurosurgeons and rehabilitation specialists, to understand attitudes toward AI-driven CDS tools before and after the 2026-2027 academic year. Members will also contribute to model development and CDS prototyping by drawing on the perspectives they study. 

2. Product prototyping with the Christensen Family Center for Innovation (CFCI) Product Lab:

Designers, product managers and engineering students will translate qualitative findings and model outputs into a prototype CDS application. The prototype will be used in usability sessions with clinicians to gather further insights and refine design features.

In the spring, the full team will integrate the model and prototype into a cohesive system and conduct provider usability tests, followed by follow-up qualitative interviews or focus groups. Surveys administered before and after the project will assess how provider attitudes toward AI change after engaging directly in tool development.

By the project’s conclusion, the team aims to produce an AI model and prototype aligned with stakeholder needs and informed by real-world workflows.

Anticipated Outputs

• Audio and video data from interviews and usability sessions
• Qualitative and quantitative analyses and related publications
• A prototype clinical decision support application
• Documentation of model development, design decisions and stakeholder-guided refinements
• Improved engagement with providers toward AI-driven CDS tools

Student Opportunities

Ideally, this project team will include 2 graduate students and 8 undergraduate students. Students from any major are welcome, and those with interests in healthcare, AI, ethics, qualitative research, and user-centered design or human-technology interaction are encouraged to apply.

The team will consist of two subgroups, qualitative and product design, supported by faculty mentors and the CFCI Product Lab. Students will gain experience in:

• Qualitative research methods, including focus group facilitation and interview-based analysis
• User-centered design, including prototype creation and usability testing
• Clinical workflow analysis and stakeholder engagement
• Interdisciplinary collaboration across healthcare, data science and design

All students will participate in weekly team meetings, journal clubs and skill-building workshops. The project includes optional Summer 2027 research opportunities to finalize analyses, publications and prototype refinements.

Timing

Fall 2026 – Summer 2027

Fall 2026:

• Conduct literature review
• Begin prototype development with CFCI Product Lab
• Plan qualitative study

Spring 2027:

• Prepare documentation for model
• Conduct qualitative study using prototype
• Integrate stakeholder feedback and iterate model
• Begin preparing publications

Summer 2027 (optional):

• Finalize model development
• Complete qualitative analysis
• Complete publications

Crediting

Academic credit available for fall and spring semesters

See earlier related team, Discovering AI Applications for Traumatic Brain Injury Care (2024-2025).