Artificial Intelligence and Ultrasonography

• Main Author: Blaivas, Michael
• Publication Date: 2024
• Format: Article
• Language: por
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: https://doi.org/10.24950/rspmi.2585
• topic: Aprendizagem Automática; Aprendizagem Profunda; Ecografia; Inteligência Artificial; Medicina Interna; Sistemas Point-of-Care; Artificial Intelligence; Deep Learning; Internal Medicine; Machine Learning; Point-of-Care Systems; Ultrasonography

Artificial intelligence (AI) and its many aliases, including machine learning, deep learning and big data, have invaded modern medicine impacting most aspects of modern practice. One of the most controversial and potentially impactful, is artificial intelligence use in medical imaging. While most commercial and academic attention has focused on higher cost imaging modalities such as magnetic imaging resonance (MRI) and computed tomography (CT), ultrasound has also become the target of AI application developers. Ultrasound presents additional barriers to AI application development and execution, not seen in axial imaging such as MRI and CT. Point-of-care ultrasound (POCUS), with its lack of standardization and plethora of inexperienced users, poses the greatest imaging challenge to AI. However, POCUS is also the key to widespread access to diagnostic and interventional ultrasound at the patient’s bedside throughout the world. This article discusses AI, it utilization in POCUS, current challenges, risks, limitations, needs and future possibilities.