A Deep Learning Technique for Multi-view Prediction of Bone

Abstract

In the medical field, day by day a new technology is introduced to reduce the efforts of doctors as well as patients. Before the actual treatment, patients' needs satisfaction to diagnose a defect in the body part. The current techniques available to detect the correct fractured/damaged bone part of a human is either a Computerized Tomography scan or Magnetic Resonance Imaging scan. The mentioned techniques are either unavailable in rural areas or are costly compare to the X-ray technique. This issue attracts the attention to design a technique that converts a 2-Dimensional (2-D) images into its equivalent 3- Dimensional (3-D) images. For this purpose, the authors used the Generative Adversarial Network to implement a technique that takes an X-ray image as input and gives its equivalent 0° to 360° images.