Spine Endoscopic Neural Segmentation Image Dataset

The Spine Endoscopic Neural Segmentation (SENDS) dataset addresses the clinical challenge of neural structure segmentation in spinal endoscopic surgery and fills the gap in publicly available annotated data. Collected from real intraoperative scenarios, the dataset presents challenging conditions such as low tissue contrast, strong specular reflections, and occlusions caused by surgical instruments. All neural regions were manually annotated by board-certified neurosurgeons, with consensus labeling applied in ambiguous areas to ensure high-quality annotation consistency. The SENDS dataset can be used for research on semantic segmentation of neural structures and is also suitable for developing and evaluating models involving super-resolution-assisted tasks.