3D Reconstruction from 2D DICOM Images with an Interactive GUI

Abstract

3D modeling is one of the most frequently used application of computer vision in the present times. The most important step in the process of 3D modeling is surface reconstruction. In the field of biomedical engineering, surface reconstruction finds an important application in orthodontics. It can be used to create 3D models of dentures using medical imaging techniques such as X-rays, MRI (Magnetic Resonance Imaging) and CT (Computed Tomography) scan as input data. These 3D models help the dentists in effective visualization, diagnosis and surgery planning. This research focuses on creating 3D models using a novel dataset and then comparing the results of reconstruction with the 3D models obtained using two freely available open-source algorithm for 3D reconstruction using the same dataset. The results are compared using both qualitative and quantitative methods. In this way, an effective 3D reconstruction algorithm is developed and tested that reconstructs efficient 3D models with the least errors.