Synthetic 3D Spinal Vertebrae Reconstruction from Biplanar X-rays Utilizing Generative Adversarial Networks

This study uses a special computer program called a Generative Adversarial Network (GAN) to create 3D models of spinal bones from 2D X-ray images. The GAN learns to combine information from two different X-ray views (front and side) to build a detailed 3D structure, similar to what you'd see in a CT scan. This method could help doctors get a better view of the spine using regular X-ray machines, which are cheaper and expose patients to less radiation than CT scans.

• 1
  The GAN-based method can effectively reconstruct 3D spinal vertebrae structures from biplanar X-rays.
• 2
  The average PSNR was 28.394 dB, PSNR-3D was 27.432, SSIM was 0.468, cosine similarity was 0.484, MAE0 was 0.034, and MAE was 85.359.
• 3
  There are limitations in accurately capturing the fine bone structures and maintaining the precise morphology of the vertebrae.