Digital Twin-Based Skill Training With a Hands-On User Interaction Device to Assist in Manual and Robotic Ultrasound Scanning

Although teleconsultation and robot-based remote teleoperation in ultrasonography have facilitated reliable image acquisition, lack of training still seriously affects the accuracy of ultrasound diagnosis. In order to improve physicians’ scanning skills, a more efficient training process is needed. However, present ultrasound training methods, such as the training phantom, are too costly for individual training. To explore new methods for training, the combined use of vision-based location tracking, motion sensing, and virtual ultrasound generation methods was studied in this work to offer a more low-cost and intuitive interaction system for virtual ultrasound scanning. The feasibility of the proposed workflow was verified by a user study on simple tasks and a clinical evaluation on a virtual ultrasound examination of three fetal heart specimens. As a result, the proposed interaction device facilitated virtual ultrasound examination by making it more efficient and intuitive, according to objective and subjective measurements. Therefore, while virtual ultrasound provides great data extensibility, the proposed device provides a smooth interaction, making the system simple and cost-effective for the end user. It is therefore concluded that the proposed method could provide a promising education system for the trainee to understand standard views of echocardiography and the corresponding anatomical correlations.