3D Reconstruction of 2D X-Ray Images

Today 2D X-Ray Images are an integral part of immediate medical centers and hospitals. They play an important role in diagnosis and treatment planning in orthopedics. This technique is globally available and is considered as fast and inexpensive in contrast to more advanced 3D imaging methods like CT (Computerised Tomography) Scan and MRI (Magnetic Resonance Imaging) Scan. The doctors use these methods to take photographs of the patient's body by placing a photographic film behind the body, with the X-Ray machine turned on. These rays take longer to pass through the bones in comparison to the skin that is why bones appear much lighter. The musculoskeletal analysis is by nature 3D, therefore, accurate assessment of a patient's anatomy is widely dependent upon a surgeon's ability to analyze the 2D images and make assumptions from the data available. To overcome the limitations of the surgeons understanding and for effective planning of surgical involvement. The goal of this project is the 3D reconstruction of images from 2D X-Ray images. Thus, in this paper, we have proposed a approach using machine learning for conversion which is independent of the experiment setup. First, Bone extraction from the image was done. Secondly, the 3D image has been obtained using stl conversion. At last the data are validated using patient data.