Standard Operating Procedures for Blood Collection Nurses: Posture and Hand Trajectory Dataset

In modern healthcare systems, the blood collection skills of nurses are a fundamental yet crucial aspect. Due to variations in operational levels among nurses, non-standard operations can affect patient experience and medical efficiency. Current training is often limited to simulated patients or theoretical teaching, lacking standardized practical operation guidance. This dataset aims to provide data support for medical training and smart robot development by recording nurses' standard postures and hand trajectories during blood collection via video. The data is collected in real hospital environments using high-frame-rate cameras, covering the actual operations of multiple professional nurses. Data quality is strictly controlled, undergoing multiple rounds of annotation and consistency checks to ensure accuracy and completeness. Annotation is performed by a professional team with nursing backgrounds, consisting of 20 people. Data preprocessing includes denoising, lens correction, video segmentation, and action classification. The final data is stored in MP4 format, organized by operational steps.