Refinement of image quality in panoramic radiography using a generative adversarial network

The study developed and evaluated a generative adversarial network (GAN) model for improving the image quality of panoramic radiography. The researchers prepared degraded image datasets using four different processing methods and trained the Pix2Pix GAN model using pairs of the original and degraded image datasets. The model was tested using the peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM), and two radiologists rated the clinical image quality. The results showed that the GAN model was effective in improving images with blur in the anterior teeth region but less effective with blur and noise. The study suggests that the developed GAN model can potentially improve panoramic radiographs with degraded image quality.