Image Denoising for Low-Dose CT via Convolutional Dictionary Learning and Neural Network

Removing noise and artifacts from low-dose computed tomography (LDCT) is a challenging task, and most existing image-based algorithms tend to blur the results. To improve the resolution of denoising results, we combine convolutional dictionary learning and convolutional neural network (CNN), and propose a transfer learning densely connected convolutional dictionary learning (TLD-CDL) framework. In detail, we first introduce the dense connections and multi-scale Inception structure to the network, and train the pre-model on the natural image dataset, then fit the model to the post-processing of LDCT images in the way of transfer learning. In addition, considering that a single pixel-level loss is difficult to achieve satisfactory results both in the index and visual perception, we use the compound loss function of L1 loss and SSIM loss to guide the training. The experimental result shows that TLD-CDL has a good balance between noise reduction and the preservation of details, and acquires inspiring effectiveness in terms of qualitative and quantitative perspective.

Automated summary

To address the issue of noise and artifacts in low-dose computed tomography (LDCT), researchers propose a transfer learning densely connected convolutional dictionary learning (TLD-CDL) framework that combines convolutional dictionary learning and convolutional neural network (CNN). By introducing dense connections and multi-scale Inception structure, the framework achieves improved resolution for denoising results. Experimental results show that TLD-CDL effectively balances noise reduction and preservation of details.