Abstract
Rapid and accurate diagnosis is critical in managing traumatic injuries both in military and civilian settings. Handheld ultrasound devices enable Point-Of-Care UltraSound (POCUS) in the field, improving triage and treatment. The extended Focused Assessment with Sonography in Trauma (eFAST) procedure is particularly useful for detecting internal bleeding, pneumothorax, and cardiac complications. However, effective use of ultrasound requires significant training and experience.
In this project, we explore how artificial intelligence (AI) can lower the threshold for using ultrasound by guiding less experienced users through the eFAST procedure. Prior studies show AI can enhance image interpretation and examination quality, particularly in identifying free fluid and pericardial effusion.
Key activities include user needs assessments, data collection, AI model training, and prototype development with a user-friendly interface. A goal is to design a system that can operate independently of cloud services and withstand cyber threats and electronic warfare.
The program has the potential to reduce unnecessary medical evacuations by improving diagnostic accuracy. By lowering the training requirements, it supports task shifting, task sharing and decision support, in line with NATO’s suggested strategy for alleviating medical workforce shortages.