Image denoising from a single-shot of OCT using deep learning

DOI

Bibliographic Information

Other Title
  • Deep learningによるシングルショットOCT画像のノイズ除去

Search this article

Description

<p>As optical coherence tomography (OCT) technology has developed, the photographing range and the image quality of fundus OCT images have improved. However, fundus OCT images contain noise such as speckle noise. Application of the averaging method will improve their image qualities. In the averaging technique, noise is reduced by scanning the same area multiple times and then averaging images to obtain a clear image. The more frames the better the image, but this requires more time for photographing, which causes a burden on patients.</p><p>We proposed noise reduction from a single scan of OCT using Deep Learning, aiming to obtain images conforming to 120 images used by the averaging method from a single-shot of OCT. The effectiveness of the proposed method is shown through both objective and subjective evaluations.</p>

Journal

Keywords

Details 詳細情報について

Report a problem

Back to top