7-DoF Laparoscopic Peg Transfer Dataset for Surgical Skill Assessment

Accurate perception of surgical instruments is crucial for automated skill assessment and computer-assisted training in minimally invasive surgery. However, publicly available video-kinematic datasets are scarce, particularly for non-in-vivo tasks. This work introduces LASK (LAparoscopic Skill & Kinematics), a peg-transfer dataset featuring synchronised HD video and 7-DoF (seven-degree-of-freedom) ground-truth kinematics for two surgical graspers. The dataset comprises 114 trials (~3 hours total) from 38 low-, 41 medium- and 35 high-skill expert surgeons, providing 324,101 frames with time-aligned kinematics for both tool and tooltips; 3,725 frames include annotated bounding boxes, and a complete 2,680-frame validation sequence. LASK distinctively captures two instruments throughout with wider fields of view than typical in-vivo data, includes surgeon-specific metadata (handedness & experience), and reflects typical box-trainer imaging conditions. These features support robust benchmarking of multi-class detection, tracking, pose estimation, and skill classification algorithms. Once publicly released, LASK aims to improve laparoscopic training by fostering data-driven training tools.