Handheld Ultrasound

Handheld ultrasound is rapidly evolving, aiming to improve accessibility and affordability of ultrasound imaging through portable devices. Current research focuses on enhancing image quality by addressing challenges like inconsistent gel application (using automated gel dispensers and deep learning for improved image processing) and achieving 3D reconstruction from freehand scans (leveraging visual-inertial odometry and deep learning-based inertial odometry). These advancements, along with training-free image style alignment techniques to adapt to variations in handheld device data, are enabling more accurate and reliable automated measurements and diagnoses in various clinical settings, including colorectal cancer detection and patellar tracking.