BACKGROUND AND OBJECTIVES:
Intraoperative orientation during microsurgery has a prolonged learning curve among neurosurgical residents. Three-dimensional (3D) understanding of anatomy can be facilitated with realistic 3D anatomic models created from photogrammetry, where a series of 2-dimensional images is converted into a 3D model. This study implements an algorithm that can create photorealistic intraoperative 3D models to exemplify important steps of the operation, operative corridors, and surgical perspectives.

METHODS:
We implemented photograph-based and video-based scanning algorithms for uptakes using the operating room (OR) microscope, targeted for superficial structures, after surgical exposure, and deep operative corridors, in cranial microsurgery. The algorithm required between 30–45 photographs (superficial scanning), 45–65 photographs (deep scanning), or approximately 1 minute of video recording of the entire operative field to create a 3D model. A multicenter approach in 3 neurosurgical departments was applied to test reproducibility and refine the method.

RESULTS:
Twenty-five 3D models were created of some of the most common neurosurgical approaches—frontolateral, pterional, retrosigmoid, frontal, and temporal craniotomy. The 3D models present important steps of the surgical approaches and allow rotation, zooming, and panning of the model, enabling visualization from different surgical perspectives. The superficial and medium depth structures were consistently presented through the 3D models, whereas scanning of the deepest structures presented some technical challenges, which were gradually overcome with refinement of the image capturing process.

CONCLUSION:
Intraoperative photogrammetry is an accessible method to create 3D educational material to show complex anatomy and demonstrate concepts of intraoperative orientation. Detailed interactive 3D models, displaying stepwise surgical case-based anatomy, can be used to help understand details of the operative corridor. Further development includes refining or automatization of image acquisition intraoperatively and evaluation of other applications of the resulting 3D models in training and surgical planning.