Feedback-Based Self-training System of Patient Transfer

In this paper, we propose a self-training system to assist nursing students to learn nursing skills. The system focuses on the task of transferring a patient from a bed to a wheelchair. In the system, two Kinect sensors were applied to measure the posture of the trainees and patients and an automatic evaluation method was used to classify the trainees' performance in each skill as correct or incorrect. A feedback interface based on a checklist was designed to help the trainees check whether they performed correctly. The system is designed for the trainees to operate by themselves. A control test was performed to measure the training effects of the system. The results show that the growth rate of the group that trained with feedback (79%) was higher than the group that trained without feedback (48%).