Emergency Medical Artificial Intelligence for LMICs
A clearer next step
after a head injury.
Our current prototype focuses on head injuries. It asks structured questions, remembers earlier calls, and calls back to check for changes, all through an ordinary phone.
Functional prototype. Not a live medical service.
1st place — mission:BRAIN Hackathon 2026OUR STARTING POINT
A case that raised a practical question.
A fictional hackathon case in rural Gilgit-Baltistan, Pakistan, described uncertainty after a head injury and a difficult journey to care. It prompted the idea behind LUCID.
The uncertaintyA person may initially appear well, while the people helping them are unsure what to look for or what to do next.
The access gapDistance and limited access to guidance can make the next step toward care harder.
Our design questionCould a familiar phone call help families recognize potential concerns and keep track of changes over time?
This case was the starting point for LUCID’s entry in the 2026 mission:BRAIN hackathon.
WHO IT HELPS
Families seeking a next step.
People with possible head injuries, and the family members or caregivers trying to help them in settings where timely guidance can be difficult to reach.
Support the person helping
The design centers on a caller who may be uncertain about the situation and needs a structured way to describe it.
Keep access familiar
An ordinary phone call, including a landline. No app or caller-side internet; the backend requires infrastructure.
PROTOTYPE WORKFLOW
From the first call to the follow-up.
Call
An ordinary telephone, including a landline. No app or internet connection on the caller’s phone.
Conversation
Conversational AI gathers context through guided questions, with explicit triage and escalation rules.
Continuity
LUCID remembers earlier answers across interrupted calls and calls back to check for changes.
Escalation
The prototype demonstrates escalation prompts and routing where configured. Local referral pathways and clinical oversight are needed before use.
Conversational AI with explicit escalation rules. Clinical review and validation are still required.
THE PROTOTYPE
A functional, private
proof of concept.
Prototype capabilitiesConversational AI, multilingual voice interaction, persistent conversation context, automated follow-up, and configured escalation or routing.
Current statusNo patient-facing service, clinical validation, or community field pilot.
The project site includes the extended YouTube recording and a separate, shortened text walkthrough with rule explanations. The recording shows the prototype; the walkthrough explains selected interactions. The phone system remains private.
LOCAL ADAPTATION
The next setting
needs its own plan.
Nigeria is the proposed next setting. LUCID and Glial Initiative are preparing an adaptation and sustainability proposal for Wema Hackaholics 7.0’s Social Impact track, shaped by local requirements.
Community needs & language
Telephone access & infrastructure
Clinical input & referral pathways
Joint operating & funding model
Language performance, telephone feasibility, and referral pathways need local assessment. Ilorin, Kwara State is the competition pool; a pilot site has not been selected.
WORKING TOGETHER
Clear contributions.
A shared operating plan.
Engineering
Retain internal control of the codebase and implement agreed prototype adaptations.
Local knowledge
Local research, community engagement, and introductions to healthcare and community stakeholders.
Sustainability
Develop the implementation, operating, and funding model jointly, including cost and affordability questions.
This is a student collaboration. Institutional hosts, funding, and live deployment commitments are not established.
NEXT STEPS & DESIRED IMPACT
Define the setting.
Then evaluate responsibly.
NowComplete the proposal, clarify local requirements, and prepare an adapted demonstration.
Before any patient useEstablish clinical ownership, language and telephone feasibility, referral pathways, consent, security, and an evaluation plan.
We hope to help families recognize potential concerns and seek appropriate help sooner.
Future evaluation could examine caller understanding, call and follow-up completion, and agreement with qualified clinical review. No impact results are claimed.
TEAM & CONTACT
Help shape the next step.
Created by a three-person University of Utah student team.
Walter Shewmake
AI architecture & engineering
David Ross
Clinical logic & decision systems
Dennis Courtright
Business strategy & concept validation
LUCID is a prototype for demonstration. This website does not provide medical advice, diagnosis, or emergency services.