Autonomous Robotic Ultrasound

Autonomous robotic ultrasound (RUS) systems aim to improve the consistency and accessibility of ultrasound imaging by automating probe manipulation and image acquisition. Current research focuses on developing algorithms for automated path planning (often using reinforcement learning or Bayesian optimization) and improving image quality through techniques like automated gel application and optimized probe orientation. These advancements leverage deep learning models, including U-Nets and convolutional neural networks, for tasks such as image segmentation and quality estimation, ultimately aiming to reduce reliance on expert sonographers and improve diagnostic accuracy and procedural efficiency in various clinical settings.