3D Reconstruction Of Anatomical Models Based On Geometric Knowledge

Geometrically accurate three-dimensional (3D) model reconstructions of anatomical shapes play a major role in medical sector, because these 3D models are currently required for clinical studies such as visualization, diagnosis, patient follow-up, pre-operative surgery planning, and bio-mechanical applications. Such 3D models are generally extracted from data sets, collected by 3D sensors, like computed tomography (CT) scan and magnetic resonance imaging (MRI). Those methods have various limitations such as high expensive, high radiation dose (for CT), requires a high volume of image data sets, and more time to generate 3D models. To overcome above limitations, and obtain accurate 3D models, many attempts have been taken based on 2D X-Ray imageries. However, the demand for automatic process to create 3D models of anatomical structures using X-ray images is still increasing. As such, the present work is an attempt for introducing a cost and time effective method to reconstruct 3D models of human bones automatically from multiple X-ray images using photogrammetric techniques.