{
  "abstract": "Traumatic injuries require rapid and accurate diagnosis to ensure effective treatment. New technologies, in the form of hand-held ultrasound devices make it possible to perform ultrasound procedures of wounded soldiers rapidly in the field. Recent studies indicate that Point-Of-Care UltraSound (POCUS) can improve diagnostic accuracy and combat casualty care and improve triage of soldiers in the field. 1 The extended Focused Assessment with Sonography in Trauma (eFAST) procedure to look for peritoneal fluid, pericardial fluid, pneumothorax and haemothorax in a trauma patient could be well suited. However, both performing an ultrasound procedure correctly and interpreting the ultrasound images is difficult and requires extensive training and experience.2 Previous studies have demonstrated the potential of artificial intelligence (AI) methods to assist with the challenges of interpreting ultrasound images. For instance, a study that examined the use of AI-guided ultrasound to assist inexperienced users in detecting free fluid in the Morrison pouch concluded that the AI helped the users achieve better quality of the examination.3 Studies have also demonstrated the detection of both free fluid in the abdomen4 and pericardial fluid5 using AI. Our hypothesis is that the threshold for using ultrasound can be significantly lowered with the help of AI.Therefore, we aim to develop an AI-guided POCUS system for the eFAST procedure to assists military healthcare providers, including those with limited ultrasound experience, in accurately diagnosing injuries such as internal bleeding, pleural effusion, cardiac tamponade, and pneumothorax (figure 1). By making ultrasound available with the help of AI in the military field, the military medical providers will be able to quickly send those with severe injuries to the correct treatment, as well as avoid evacuating soldiers when it is not necessary. This will reduce training requirements for military healthcare providers, improve triage and treatment efficiency in combat scenarios, and align with NATO’s focus on task shifting and clinical decision support for non-specialists, addressing critical needs in military prehospital emergency care.The project involves collaboration between the research institute SINTEF Digital, St. Olav’s University Hospital in Trondheim, Norway, and the Norwegian Armed Forces Joint Medical Services. Key activities include iterative user needs assessments, data collection from the emergency department at St. Olavs University Hospital, training of modern AI models for real-time image segmentation, detection and guiding, and creation of a prototype including a graphical user interface for the AI-guided eFAST procedure. The project also emphasizes the importance of situational adaptability, ensuring that the technology can function in the field without reliance on cloud services and withstand cyber threats.The project is expected to contribute to a sustainable and more efficient military prehospital emergency care by reducing unnecessary patient evacuation and optimizing resource utilization. Ethical considerations, including data privacy and responsible innovation, are integral to the project’s execution. The collaboration aims to generate new socio-technical knowledge and practical insights into the deployment of AI-guided POCUS in military settings, with potential applications in civilian emergency care.Abstract A03 Figure 1Illustration of the concept where AI-based automatic image interpretation and guidance of the eFAST procedure can help the user. (Illustration partly made with help from ChatGPT)References Savell SC, Baldwin DS, Blessing A, Medelllin KL, Savell CB, Maddry JK. Military Use of Point of Care Ultrasound (POCUS). J Spec Oper Med. 2021 Summer;21(2):35–42.Blehar DJ, Barton B, Gaspari RJ. Learning curves in emergency ultrasound education. Acad Emerg Med. 2015 May;22(5):574–82.Chiu IM, Lin CR, Yau FF, Cheng FJ, Pan HY, Lin XH, Cheng CY. Use of a deep-learning algorithm to guide novices in performing focused assessment with sonography in trauma. JAMA Netw Open. 2023 Mar 1;6(3):e235102..Lin Z, Li Z, Cao P, Lin Y, Liang F, He J, Huang L. Deep learning for emergency ascites diagnosis using ultrasonography images. J Appl Clin Med Phys. 2022 Jul;23(7):e13695.Yıldız Potter İ, Leo MM, Vaziri A, Feldman JA. Automated detection and localization of pericardial effusion from point-of-care cardiac ultrasound examination. Med Biol Eng Comput. 2023 Aug;61(8):1947–1959.Conflict of Interest Statement The authors declare that there are no conflicts of interest regarding this project or the publication of this paper.Project funding The project is funded by The Research Council of Norway, project number 354740.",
  "authors": [
    {
      "affiliations": [
        "Department of Health Research, SINTEF Digital, Trondheim, Norway"
      ],
      "name": "Cecilie Våpenstad"
    },
    {
      "affiliations": [
        "Department of Health Research, SINTEF Digital, Trondheim, Norway"
      ],
      "name": "Sigrid Berg"
    },
    {
      "affiliations": [
        "Department of Health Research, SINTEF Digital, Trondheim, Norway"
      ],
      "name": "Mathilde Gajda Faanes"
    },
    {
      "affiliations": [
        "Department of Health Research, SINTEF Digital, Trondheim, Norway"
      ],
      "name": "Lars Eirik Bø"
    },
    {
      "affiliations": [
        "Department of Radiology and Nuclear Medicine, St. Olav’s University Hospital, Trondheim, Norway"
      ],
      "name": "Mathias Leth Kristensen"
    },
    {
      "affiliations": [
        "Norwegian Armed Forces Joint Medical Services, Sessvollmoen, Norway",
        "Department of Traumatology, St. Olav’s University Hospital, Trondheim, Norway"
      ],
      "name": "Christian Medby"
    }
  ],
  "title": "A03 Artificial intelligence-guided point-of-care ultrasound for combat casualty care",
  "uid": "2a92d497-3109-59d2-ab69-14c5403ff86c"
}