Low-dose CT image denoising method based on multi-scale residual error coding and decoding network

• Main Authors: GAO TIANHAN, WU CHANGHE
• Format: Patent
• Language: Chinese; English
• Published: 25.07.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The invention provides a low-dose CT image denoising method based on a multi-scale residual error coding and decoding network, and relates to the technical field of medical image processing. The method comprises the following steps: acquiring a data set, and selecting the data set publicly used in a low-dose CT image denoising contest of the Meio medical center; building a multi-scale residual error coding and decoding network model, and initializing network parameters to obtain an initialized multi-scale residual error coding and decoding network; inputting a to-be-processed CT image into the initialized multi-scale residual error coding and decoding network model, selecting an Adam optimizer, and training according to an error calculated by an MSE loss function to obtain a trained multi-scale residual error coding and decoding network model; and denoising the low-dose CT image by using the trained multi-scale residual coding and decoding network model to obtain a final denoised image. Noise and artifacts in