Design and development of virtual scenes using 3D models for Computer-Assisted Learning

Abstract

This work creates a 3D anatomical models of the brain as well as its virtual images. It is to be applied in virtual training of neuroanatomy. These models are based on studies of Magnetic Resonance (MRI) and Computer Tomography (CT). These models contain information of the anatomical structures of the brain's soft tissues. This research was developed using techniques of virtual training. Virtual training has become a new teaching environment for learning complex disciplines, ideal for the Medical Sciences. Virtual-Reality training has a huge impact because of its versatility of interaction with the 3D models. The students follow their own learning pace using this technique. The following method was carried out based on image-studies obtained by MRI and CT. This method was developed on three steps (1) one process for image filtering, (2) one process for image segmentation with the objective of obtaining the brain's structures, and (3) one process for creation of the 3D models. (The models were designed using the 3D Slicer program). Subsequently, the scene which includes the anatomical models were created. These scenes are achieved by applying several effects such as transparency, clipping, fusion, and registration with the original slices of this study. As a result, the virtual images are created. The students have the possibility to visualize and manipulate them, in different ways using the virtual brain structures. These 3D models could be clipped by the user or be shown jointed with the original 2D slice of the MRI or CT. These images could be exported to VRML format, which can be read using one plug-in. This research is destined to achieving a virtual environment for neurosurgical training that could be applied with minimal invasive neurosurgery. © 2009 Springer Berlin Heidelberg.