Introduction: The integration of AI into the field of disaster medicine represents a significant advancement in improving the speed and efficiency of emergency response by facilitating faster decision-making processes, optimizing resource allocation, and enhancing communication systems. Given the effectiveness of AI, it is essential to explore its potential applications within disaster medicine. To investigate this topic, we conducted a scoping review on AI applications aimed at enhancing the efficiency of emergency response and minimizing the time to care. Methods: Following the PRISMA-ScR guidelines, a comprehensive search was conducted across five databases. Peer-reviewed studies published in English between 2019 and 2024 that focused on AI applications in disaster medicine were included. Two reviewers independently screened the titles, abstracts, and full texts. Relevant findings were extracted in a thematic format. Results: Out of 15,431 studies, 239 were included. The volume of publications increased from 8 in 2019 to 59 in 2021. Research output was predominantly concentrated in Asia (40%), followed by multi-country collaborations (21%), with both Europe and North America contributing (15%) each. The primary focus of the studies was on machine learning (65%), often in combination with deep learning (50%) and decision-support systems (36%). Common applications identified included decision-support tools (86%), AI-assisted triage (63%), early warning systems (52%), and patient monitoring (49%). Notably, most of the studies (84%) concentrated on pandemics, while natural disasters were (16%), and humanitarian crises (3%). Conclusion: The results of our research indicate that the COVID-19 pandemic has resulted in rapid advancements in innovative applications, primarily focused on enhancing situational awareness and optimizing triage procedures in pandemic-related emergencies. To evaluate the life-saving potential of AI, it is essential to integrate these technologies into a wider range of disaster scenarios, supported by rigorous validation processes. This approach will improve the ability of responders to protect communities during future disasters.
Masri et al. (Sun,) studied this question.