Background Neurosurgical training usually requires long hours for hands-on procedures, making it difficult for inexperienced surgeons to quickly learn in an error-proof environment. The objective of this study was to propose a puzzle-like new model for neurosurgical education that simulates craniosynostosis correction (scaphocephaly type) using Renier's H technique. A model of a 3-dimensional (3D) anatomic simulator for craniosynostosis training is presented and evaluated. Methods The cranial model was created using 1-mm computed tomography scan images from patients with scaphocephaly in the Digital Imaging and Communications in Medicine format. This information was processed using an algorithm to generate a 3D biomodel in resin. The puzzle model and its variable training models were assessed qualitatively by a team of expert neurosurgeons. Next, the model was applied in trainees and was evaluated using specific questionnaires. Results Experts and trainees evaluated the model. The mean number of attempts without errors was 2.3 ± 0.675, for 1 error was 2.2 ± 0.918, and for 2 errors was 1.3 ± 0.707. The mean score of the simulator was 9.2 ± 0.421. Twelve residents (second evaluation) answered the questionnaire with a positive assessment of diagnosis capabilities, appropriateness of the model, time commitment, adequate environment, reliable 3D reconstruction, and teaching method. Three participants had used a 3D simulator previously, and the simulator was evaluated obtaining a 9.9 final average (range, 0–10). Conclusions The puzzle may be a complementary tool for surgical training. It allows several degrees of immersion and realism, offering symbolic, geometric, and dynamic information with 3D visualization. It provides additional data to support the practice of complex surgical procedures without exposing real patients to undue risk.
