Real-time trajectory prediction of laparoscopic instrument tip based on long short-term memory neural network in laparoscopic surgery training.

Background: To provide appropriate surgical training guidance, some skill evaluation and safety detection methods have been developed. However, these methods are difficult to provide predictive information for trainees. This paper proposes a new approach for real-time trajectory prediction of the laparoscopic instrument tip to improve surgical training and the patient safety.