Ultrasound Nerve Segmentation using YOLO

Ultrasonic neural division is a medical activity undertaken to diagnose nerve block procedures and schedule operations. Manual segmentation is labour-intensive, highly subjective and error-prone due to the complicated structure of nerves, ultrasound interference, and lack of contrast and complexity. Manual segmentation has a number of limitations but the emergence of deep learning has made fully automatic segmentation methods possible. Advances made in deep-learning architectures, and specifically U-Net and U-Net++, which can utilize skip-nested connections, have improved feature learning. YOLO architecture also allows for real-time object detection increasing clinical usage, minimizing manual workloads, improving accuracy, and maximum reliability of identifying nerves.