2D-to-3D: A Review for Computational 3D Image Reconstruction from X-ray Images

In the clinical research, three-dimensional/volumetric anatomical structure of the human body is very significant for diagnosis, computer-aided surgery, surgical planning, patient follow-up, and biomechanical applications. Medical imaging procedures including MRI (Magnetic Resonance Imaging), CT (Computed Tomography), and CBCT (Cone-beam computed tomography) have certain drawbacks such as radiation exposure, availability, and cost. As a result, 3D reconstruction from 2D X-ray images is an alternative way of achieving 3D models with significantly low radiation exposure to the patient. The purpose of this study is to provide a comprehensive view of 3D image reconstruction methods using X-ray images, and their applicability in the various anatomical sections of the human body. This study provides a critical analysis of the computational methods, requirements and steps for 3D reconstruction. This work includes a comparative critical analysis of the state-of-the-art approaches including the feature selection along with their benefits and drawbacks. This review motivates the researchers to work for 3D reconstruction using X-ray images as only a limited work is available in the area. It may provide a solution for many experts who are looking for techniques to reconstruct 3D models from X-ray images for clinical purposes.