CTformer: convolution-free Token2Token dilated vision transformer for low-dose CT denoising

Wang, Dayang; Fan, Fenglei; Wu, Zhan; Li, Rui; Wang, Fei; Yu, Hengyong

doi:10.1088/1361-6560/acc000

Cited by 81 publications

(18 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the experiment, the CT images of 9 patients (760 pairs) were used as the training set, and the images of 1 patient (35 pairs) were used as the test set. To verify further the effectiveness of the method in this study, we performed a comparison of the proposed network with seven denoising networks, namely REDCNN, 14 EDCNN, 28 wavelet- and CNN-based wavelet domain residual network (WavResNet), 37 GAN network-based WGAN_VGG, 19 Pix2Pix, 38 CNCL, 22 and CTformer 23 for visual effectiveness and quantitative evaluation. The quantitative metrics include SSIM 39 based on structural differences, peak signal-to-noise ratio (PSNR) based on pixel gray scale differences, gradient magnitude similarity deviation (GMSD) 40 based on gradient variation, feature similarity index measure (FSIM) 41 based on feature differences, and multiscale pixel domain implementation (VIFS) 42 based on visual perception.…”

Section: Resultsmentioning

confidence: 99%

“…The network has strong generalization ability and has good performance on CT, magnetic resonance, and positron emission tomography images. Wang et al 23 . proposed a pure transformer CTformer for LDCT denoising, which was a pioneer in applying visual transformer to LDCT denoising problems and introduced expansion and cyclic displacement to capture longer-range interactions.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Network combining edge features and pyramid structure for low-dose computed tomography denoising

Zhang

Liu

et al. 2023

J. Electron. Imag.

View full text Add to dashboard Cite

To balance noise reduction and texture details of low-dose computed tomography (LDCT) images during deep learning training, we propose a network combining edge features and a pyramid structure for LDCT denoising (PLEDNet). PLEDNet contains three states: superficial feature acquisition, deep feature acquisition, and feature confluence. First, the superficial feature acquisition stage uses a simple convolutional block to perform feature extraction on the input LDCT image. Second, in the deep feature extraction phase, one branch uses pyramid pooling to extract multiscale features, and the other uses the trainable Laplacian of the Gaussian module to extract edge detail features. Finally, the feature fusion phase uses a fusion attention module to adaptively fuse multiscale and edge features. The results of the experiments demonstrate that the method not only enhances various relative objective indicators but also effectively represses the noise in LDCT images while fully conserving the image texture details, which has great advantages in terms of visual effects.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Network combining edge features and pyramid structure for low-dose computed tomography denoising

Zhang

Liu

et al. 2023

J. Electron. Imag.

View full text Add to dashboard Cite

show abstract

“…It is known that, the core component of Transformer model is the self-attention module which can capture the global dependences of patches by conducting message passing on different patches to obtain more contextual representations. For example, Wang et al (2023a) propose a Convolution-free Token2Token Dilated Vision Transformer (CTformer) based on more powerful token rearrangement to encompass local contextual information. Wang et al (2021a) propose a learning model based on triple attention on both spatial and channel dimensions.…”

Section: Introductionmentioning

confidence: 99%

Chest x-ray diagnosis via spatial-channel high-order attention representation learning

Gao,

Jiang,

Wang

et al. 2024

Phys. Med. Biol.

View full text Add to dashboard Cite

Objective. Chest x-ray image representation and learning is an important problem in computer-aided diagnostic area. Existing methods usually adopt CNN or Transformers for feature representation learning and focus on learning effective representations for chest x-ray images. Although good performance can be obtained, however, these works are still limited mainly due to the ignorance of mining the correlations of channels and pay little attention on the local context-aware feature representation of chest x-ray image. Approach. To address these problems, in this paper, we propose a novel spatial-channel high-order attention model (SCHA) for chest x-ray image representation and diagnosis. The proposed network architecture mainly contains three modules, i.e. CEBN, SHAM and CHAM. To be specific, firstly, we introduce a context-enhanced backbone network by employing multi-head self-attention to extract initial features for the input chest x-ray images. Then, we develop a novel SCHA which contains both spatial and channel high-order attention learning branches. For the spatial branch, we develop a novel local biased self-attention mechanism which can capture both local and long-range global dependences of positions to learn rich context-aware representation. For the channel branch, we employ Brownian Distance Covariance to encode the correlation information of channels and regard it as the image representation. Finally, the two learning branches are integrated together for the final multi-label diagnosis classification and prediction. Main results. Experiments on the commonly used datasets including ChestX-ray14 and CheXpert demonstrate that our proposed SCHA approach can obtain better performance when comparing many related approaches. Significance. This study obtains a more discriminative method for chest x-ray classification and provides a technique for computer-aided diagnosis.

show abstract

“…2 Common to all of these in vivo multi-channel imaging applications is the need to manage ionizing radiation dose applied to the subject. In the past, iterative reconstruction techniques, including regularization based on prior assumptions, were the preeminent means of reducing dose and associated image noise in CT. More recently, however, there has been an explosion of interest in supervised deep learning (DL) methods for CT image noise removal and dose management [3][4][5] and for the augmentation of iterative reconstruction. [6][7][8] Supervised DL methods offer several key advantages over classic denoising methods and iterative reconstruction techniques.…”

Section: Introductionmentioning

confidence: 99%

Unsupervised learning of robust models for cardiac and photon-counting x-ray CT denoising

Clark

Schwartz

Euler

et al. 2023

Medical Imaging 2023: Physics of Medical Imaging

View full text Add to dashboard Cite

Supervised deep learning methods have rapidly advanced to the forefront of medical imaging research; however, they face limitations in advanced CT applications. For instance, label generation is difficult for retrospective cardiac CT data where dose modulation yields variable image quality by phase. Similarly, a model trained to denoise multi-energy data may not generalize across differing contrast and energy channel counts. In this work, we propose and demonstrate several innovations for unsupervised denoising of such multi-channel CT data sets with a focus on improving model robustness to varying noise levels, image contrast, and channel counts. Specifically, we update our past Constrained Bregman Framework for unsupervised denoising to simplify model training procedures, to automate regularization hyperparameter selection, and to spatially adapt denoising performance. Furthermore, we propose a new data normalization strategy involving weighted singular value decomposition, convolutional mean subtraction, and noise variance scaling which abstracts the image denoising problem from the contrast and channel count of the input data. We combine these improvements with a hybrid Inception Net, half U-Net network structure, ideally suited for 3D data processing, and a plug-and- play scheme for incorporating traditional regularizers into deep learning cost functions. We demonstrate the value of these improvements for network training and denoising using retrospectively gated cardiac photon-counting CT data acquired with a Siemens NAEOTOM Alpha scanner (2-3x noise reduction). This clinical model robustly generalizes to 40-channel preclinical cardiac, photon-counting CT data acquired in a mouse without the need for additional training (6- 10x noise reduction, negligible intensity bias).

show abstract

CTformer: convolution-free Token2Token dilated vision transformer for low-dose CT denoising

Cited by 81 publications

References 56 publications

Network combining edge features and pyramid structure for low-dose computed tomography denoising

Network combining edge features and pyramid structure for low-dose computed tomography denoising

Chest x-ray diagnosis via spatial-channel high-order attention representation learning

Unsupervised learning of robust models for cardiac and photon-counting x-ray CT denoising

Contact Info

Product

Resources

About