Integrating Wearable Sensor Data With an AI-Based, Protocol-Flexible Triage Platform to Accelerate Decision-Making During the Golden Hour of Combat Casualty Care
Introduction Rapid, accurate triage within the first "golden hour" on the battlefield is critical to survival, yet existing wearable sensors and manual protocols operate in isolation, delaying care and creating inconsistent priority ratings across military and civilian systems. We therefor...
Uloženo v:
| Vydáno v: | Curēus (Palo Alto, CA) Ročník 17; číslo 8; s. e91121 |
|---|---|
| Hlavní autor: | |
| Médium: | Journal Article |
| Jazyk: | angličtina |
| Vydáno: |
United States
01.08.2025
|
| Témata: | |
| ISSN: | 2168-8184, 2168-8184 |
| On-line přístup: | Získat plný text |
| Tagy: |
Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
|
| Shrnutí: | Introduction Rapid, accurate triage within the first "golden hour" on the battlefield is critical to survival, yet existing wearable sensors and manual protocols operate in isolation, delaying care and creating inconsistent priority ratings across military and civilian systems. We therefore designed a vendor-agnostic, protocol-flexible platform that ingests live biometric streams from military-grade wearables, applies an artificial intelligence (AI)-enhanced hybrid decision engine, and returns colour-coded casualty categories under any standard (e.g., MARCH (Massive hemorrhage, Airway, Respiration, Circulation, and Hypothermia/Head injury), Canadian Triage and Acuity Scale (CTAS), START(Screening Tool to Alert to the Right Treatment), SALT (Sort, Assess, Life-saving Interventions, Treatment/Transport)) in real time. Methods The planned platform will comprise four layers: (i) Data Ingestion, which normalizes ECG, heart-rate, blood-pressure, and oxygen saturation (SpO₂) signals to Open mHealth/Fast Healthcare Interoperability Resources (FHIR) schemas via BLE (Bluetooth Low Energy) or WebSockets; (ii) Secure Transport, which encrypts packets with Transport Layer Security (TLS) 1.3 and uses store-and-forward buffering through message queuing telemetry transport (MQTT)/Kafka brokers; (iii) Triage Engine, which merges continuous machine-learning severity scores with dynamically loaded guideline definitions (JSON (JavaScript Object Notation)/YAML (YAML Ain't Markup Language)) and rule-based overrides; and (iv) Presentation, which delivers one-tap guideline selection and actionable prompts to medics while providing a geo-tagged command dashboard. Validation will combine bench tests on synthetic and de-identified Medical Information Mart for Intensive Care IV (MIMIC-IV) datasets with randomized field-simulation trials that compare manual versus AI-augmented triage, measuring time to triage, protocol concordance, and medic cognitive load (National Aeronautics and Space Administration Task Load Index(NASA-TLX)). Results (projected) We anticipate at least a 30% reduction in triage time, agreement rates of 85% or higher with expert assessments across multiple protocols, and significant decreases in medic workload. Conclusion A dynamic, standards-based artificial intelligence (AI) triage platform could harmonize military and civilian casualty care, accelerate golden-hour decision-making, and improve multinational interoperability. Planned live-exercise evaluations and open-source protocol libraries are expected to facilitate rapid adoption and continuous refinement. |
|---|---|
| Bibliografie: | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23 |
| ISSN: | 2168-8184 2168-8184 |
| DOI: | 10.7759/cureus.91121 |