Search

Background: Recent technological improvements in three-dimensional (3D) virtual modeling software have allowed for the generation of digital anatomical models from non-contrast computed tomography (CT) images. In this study, high‐resolution non-contrast CT images of an adult female were used to develop a novel, clinically relevant digital model of the mediastinum for medical education. This model was designed to teach students complex relationships within the mediastinum that may be otherwise difficult to appreciate in traditional cadaveric specimens. Methods: A unique adaptation of semi-auto combined rendering was used to create the model, improving upon previously used rendering techniques. Post-segmentation, a non-Gaussian 3D smoothing technique was applied to maintain anatomical volumetric and boundary relationships between anatomical structures within the model. Unlike many current models, the mediastinum model was developed with attention to cognitive load theory. Results: The resulting model was intentionally colored to reduce extraneous cognitive load and enhance retention of anatomical and spatial relationships. A short video of the animated model was created to demonstrate key views and relationships in a way that would maximize learning. Conclusions: The application of semi-auto combined rendering provided an efficient way to create a digital mediastinum model while optimizing anatomical accuracy. This novel 3D digital mediastinum model allows for easy visualization, manipulation, and interaction on a computer, and can be presented in a stereoscopic 3D virtual environment to optimize learner retention.