End-to-end deep learning for interior tomography with low-dose x-ray CT

Abstract

Objective: There exist several X-ray computed tomography (CT) scanning strategies to reduce a radiation dose, such as (1) sparse-view CT, (2) low-dose CT, and (3) region-of-interest (ROI) CT (called interior tomography). To further reduce the dose, the sparse-view and/or low-dose CT settings can be applied together with interior tomography. Interior tomography has various advantages in terms of reducing the number of detectors and decreasing the X-ray radiation dose. However, a large patient or small field-of-view (FOV) detector can cause truncated projections, and then the reconstructed images suffer from severe cupping artifacts. In addition, although the low-dose CT can reduce the radiation exposure dose, analytic reconstruction algorithms produce image noise. Recently, many researchers have utilized image- domain deep learning (DL) approaches to remove each artifact and demonstrated impressive performances, and the theory of deep convolutional framelets supports the reason for the performance improvement. Approach: In this paper, we found that the image- domain convolutional neural network (CNN) is difficult to solve coupled artifacts, based on deep convolutional framelets. Significance: To address the coupled problem, we decouple it into two sub-problems: (i) image-domain noise reduction inside truncated projection to solve low-dose CT problem and (ii) extrapolation of projection outside truncated projection to solve the ROI CT problem. The decoupled sub-problems are solved directly with a novel proposed end-to-end learning using dual-domain CNNs. Main results: We demonstrate that the proposed method outperforms the conventional image-domain deep learning methods, and a projection-domain CNN shows better performance than the image-domain CNNs which are commonly used by many researchers.

An Artificial Neural Network-based Radiomics model for Predicting Radiotherapy response of Advanced Esophageal Squamous Cell Carcinoma patients: A multi-center Study

“…Screening the reproducibility of features, which is considered an indispensable part of reducing the over tting of traditional machine learning radiomic models, seems to improve the performance of the radiomics model. In contrast, the end-to-end algorithms do not need to reduce the dimensionality and aim to make full use of all image information to draw conclusions [51] . However, the superiority of these two types algorithms have not been compared to predict radiotherapy response.…”

Dose reduction and image enhancement in micro‐CT using deep learning

“…[14][15][16] Note that, besides image-to-image DL frameworks that serve as post-reconstruction tool, alternatives are to train the DL denoising model in the projection space or to develop an end-to-end DL algorithm that directly maps a noisy sinogram to a clean image (DL-based image reconstruction). 29,30 While the principles of DL denoising are equally applicable to clinical and preclinical settings, most research efforts are spent on clinical low dose CT and only a few studies investigate the merits of DL-based image enhancement for preclinical imaging. Chen et al 31 proposed a conditional GAN as framework for denoising micro-CT in the projection domain, and low dose micro-CT scans were artificially created by adding Poisson noise into the high dose projections.…”

“…For clinical low dose CT, a multitude of DL denoising models have been proposed, and each investigates different combinations of network architectures and loss functions 14–16 . Note that, besides image‐to‐image DL frameworks that serve as post‐reconstruction tool, alternatives are to train the DL denoising model in the projection space or to develop an end‐to‐end DL algorithm that directly maps a noisy sinogram to a clean image (DL‐based image reconstruction) 29,30 …”

Ring artifacts suppression for X-ray CT images by fusion of dual-domain images based on improved UNet

“…Relevant personnel have tried to use convolutional neural networks to process CT images, and have achieved certain results. For example, Chen, Usui, and Han used the convolutional neural network model to greatly reduce the content of noise in the image for the problem of the low signal-to-noise ratio of low-dose CT images [20][21][22]. Junghyun and Haofu respectively constructed neural network models to suppress metal artifacts in CT images [23,24].…”