Expert Consensus on Data Annotation and Quality Control for Trajectory Planning of Thoracolumbar Pedicle Screw Placement

In recent years, surgical robot systems for pedicle screw placement developed with artificial intelligence (AI) technologies have demonstrated substantial potential, offering enhanced accuracy and efficiency in preoperative trajectory planning while optimizing surgical safety. Such AI-assisted planning systems can precisely calculate the entry point, endpoint, and trajectory of screws, effectively avoiding damage to critical anatomical structures and assisting surgeons in developing rational surgical strategies. However, the performance and clinical applicability of these algorithmic systems rely heavily on datasets with high-quality annotated results. To standardize the construction of related datasets, the National Institutes for Food and Drug Control (NIFDC), in collaboration with relevant institutions, has developed this expert consensus, aiming to provide standardized guidance for the establishment of pedicle screw placement trajectory planning datasets.