INTRODUCTION
"Creation of a Mobile Stroke Detection and Assessment Tool Using Smartphone Sensors" is a perfect LMES STEM capstone topic—it directly addresses medical readiness, telemedicine in austere environments, and rapid response technologies for expeditionary forces. This has immediate relevance for shipboard medicine, remote base clinics, and field emergency response.
Operational: F.A.S.T.-Track: A Smartphone-Based Pre-Hospital Stroke Assessment System for Remote Military Medics
Technical: Multi-Modal Sensor Fusion for Stroke Symptom Detection Using Inertial Measurement Units and Facial Analysis
Clinical: A Mobile Decision Support Tool for Stroke Triage Using Computer Vision and Motion Analysis
Operational Gap: Stroke is time-critical ("time is brain"). In remote deployments, neurological expertise is unavailable, and evacuation windows are narrow.
Force Readiness: Strokes can occur in younger personnel (due to trauma, emboli, or undiagnosed conditions). Rapid identification preserves force strength and veteran outcomes.
Technology Alignment: Leverages ubiquitous hardware (smartphones) already in the inventory—no specialized medical device logistics required.
Dual-Use Potential: Direct application to civilian EMS, rural healthcare, and veteran home monitoring.
Activity: "Paper Prototyping & User Flow"
Paper
Markers
sticky notes
"smartphone frame" printouts
tape/glue
Scissors will be used under the facilitator’s guidance.
Prototype Application Development
Platform: React Native or Flutter for cross-platform deployment, or native Android for deeper sensor access.
User Flow:
Medic opens app, selects "Stroke Assessment"
Guided, timed tests (60 seconds total):
"Look at the camera and smile"
"Hold your arms out for 10 seconds" (phone placed on forearm)
"Repeat this phrase: 'You can't teach an old dog new tricks'"
Immediate result: "Low/Medium/High probability of stroke. Recommend urgent tele-neurology consult."
Critical Feature: Offline-first operation (essential for shipboard/remote use with limited connectivity).
Innovation Pathways (For Higher Complexity)
Multi-Person Assessment: Use LiDAR/ToF sensors (on newer phones) for 3D motion capture of gait and arm movements.
Continuous Monitoring: For high-risk patients (e.g., post-cardiac procedure), implement background monitoring of speech patterns during normal phone use.
Embedded AI: Use TensorFlow Lite or Core ML to run models entirely on-device, ensuring data privacy and offline operation.
Integration with Medical Kits: Design a 3D-printed mount to secure phone to patient's arm for standardized motor testing.
Deliverables for Your Capstone
Functional Prototype App: Deployable on at least one platform (Android/iOS).
Validation Report: Performance metrics (sensitivity, specificity) against simulated cases.
Technical Documentation: Sensor fusion algorithm, architecture decisions, code repository.
LMES Integration Paper: How this tool fits into existing military medical protocols (e.g., TCCC, shipboard sick call).
Why This is a Strong LMES Topic
Direct Impact: Addresses a time-sensitive, high-acuity medical emergency common in aging populations (including senior officers and veterans).
Resourcefulness: Uses existing hardware (every medic has a smartphone) rather than requiring new equipment.
Systems Thinking: Involves sensor integration, data fusion, human-computer interaction, and clinical workflow design.
Portfolio-Ready: Results in a tangible, demonstrable app that shows both technical and operational thinking.
Next Step: If you pursue this, immediately seek collaboration with:
A clinical advisor (nurse practitioner, EMT instructor, military medic)
Human subjects approval for your validation study (through your university's IRB)
Potential DoD partnership through your institution's ROTC or research office
This project has the potential to transition from a capstone to a funded SBIR topic or startup venture, given its clear dual-use potential and alignment with military medical innovation priorities
Posts
