CT Noise-Reduction Methods for Lower-Dose Scanning: Strengths and Weaknesses of Iterative Reconstruction Algorithms and New Techniques.

Iterative reconstruction (IR) algorithms are the most widely used CT noise-reduction method to improve image quality and have greatly facilitated radiation dose reduction within the radiology community. Various IR methods have different strengths and limitations. Because IR algorithms are typically nonlinear, they can modify spatial resolution and image noise texture in different regions of the CT image; hence traditional image-quality metrics are not appropriate to assess the ability of IR to preserve diagnostic accuracy, especially for low-contrast diagnostic tasks. In this review, the authors highlight emerging IR algorithms and CT noise-reduction techniques and summarize how these techniques can be evaluated to help determine the appropriate radiation dose levels for different diagnostic tasks in CT. In addition to advanced IR techniques, we describe novel CT noise-reduction methods based on convolutional neural networks (CNNs). CNN-based noise-reduction techniques may offer the ability to reduce image noise while maintaining high levels of image detail but may have unique drawbacks. Other novel CT noise-reduction methods are being developed to leverage spatial and/or spectral redundancy in multiphase or multienergy CT. Radiologists and medical physicists should be familiar with these different alternatives to adapt available CT technology for different diagnostic tasks. The scope of this article is (a) to review the clinical applications of IR algorithms as well as their strengths, weaknesses, and methods of assessment and (b) to explore new CT image reconstruction and noise-reduction techniques that promise to facilitate radiation dose reduction. ©RSNA, 2021.

Evaluating a Convolutional Neural Network Noise Reduction Method When Applied to CT Images Reconstructed Differently Than Training Data.

OBJECTIVE: The aim of this study was to evaluate a narrowly trained convolutional neural network (CNN) denoising algorithm when applied to images reconstructed differently than training data set. METHODS: A residual CNN was trained using 10 noise inserted examinations. Training images were reconstructed with 275 mm of field of view (FOV), medium smooth kernel (D30), and 3 mm of thickness. Six examinations were reserved for testing; these were reconstructed with 100 to 450 mm of FOV, smooth to sharp kernels, and 1 to 5 mm of thickness. RESULTS: When test and training reconstruction settings were not matched, there was either reduced denoising efficiency or resolution degradation. Denoising efficiency was reduced when FOV was decreased or a smoother kernel was used. Resolution loss occurred when the network was applied to an increased FOV, sharper kernel, or decreased image thickness. CONCLUSIONS: The CNN denoising performance was degraded with variations in FOV, kernel, or decreased thickness. Denoising performance was not affected by increased thickness.

Multiple Kernel Synthesis of Head CT Using a Task-Based Loss Function.

In CT imaging of the head, multiple image series are routinely reconstructed with different kernels and slice thicknesses. Reviewing the redundant information is an inefficient process for radiologists. We address this issue with a convolutional neural network (CNN)-based technique, synthesiZed Improved Resolution and Concurrent nOise reductioN (ZIRCON), that creates a single, thin, low-noise series that combines the favorable features from smooth and sharp head kernels. ZIRCON uses a CNN model with an autoencoder U-Net architecture that accepts two input channels (smooth- and sharp-kernel CT images) and combines their salient features to produce a single CT image. Image quality requirements are built into a task-based loss function with a smooth and sharp loss terms specific to anatomical regions. The model is trained using supervised learning with paired routine-dose clinical non-contrast head CT images as training targets and simulated low-dose (25%) images as training inputs. One hundred unique de-identified clinical exams were used for training, ten for validation, and ten for testing. Visual comparisons and contrast measurements of ZIRCON revealed that thinner slices and the smooth-kernel loss function improved gray-white matter contrast. Combined with lower noise, this increased visibility of small soft-tissue features that would be otherwise impaired by partial volume averaging or noise. Line profile analysis showed that ZIRCON images largely retained sharpness compared to the sharp-kernel input images. ZIRCON combined desirable image quality properties of both smooth and sharp input kernels into a single, thin, low-noise series suitable for both brain and skull imaging.

A Classification-Based Adaptive Segmentation Pipeline: Feasibility Study Using Polycystic Liver Disease and Metastases from Colorectal Cancer CT Images.

Automated segmentation tools often encounter accuracy and adaptability issues when applied to images of different pathology. The purpose of this study is to explore the feasibility of building a workflow to efficiently route images to specifically trained segmentation models. By implementing a deep learning classifier to automatically classify the images and route them to appropriate segmentation models, we hope that our workflow can segment the images with different pathology accurately. The data we used in this study are 350 CT images from patients affected by polycystic liver disease and 350 CT images from patients presenting with liver metastases from colorectal cancer. All images had the liver manually segmented by trained imaging analysts. Our proposed adaptive segmentation workflow achieved a statistically significant improvement for the task of total liver segmentation compared to the generic single-segmentation model (non-parametric Wilcoxon signed rank test, n = 100, p-value << 0.001). This approach is applicable in a wide range of scenarios and should prove useful in clinical implementations of segmentation pipelines.

Supervised Learning for CT Denoising and Deconvolution Without High-Resolution Reference Images.

Purpose: Convolutional neural networks (CNNs) have been proposed for super-resolution in CT, but training of CNNs requires high-resolution reference data. Higher spatial resolution can also be achieved using deconvolution, but conventional deconvolution approaches amplify noise. We develop a CNN that mitigates increasing noise and that does not require higher-resolution reference images. Methods: Our model includes a noise reduction CNN and a deconvolution CNN that are separately trained. The noise reduction CNN is a U-Net, similar to other noise reduction CNNs found in the literature. The deconvolution CNN uses an autoencoder, where the decoder is fixed and provided as a hyperparameter that represents the system point spread function. The encoder is trained to provide a deconvolution that does not amplify noise. Ringing can occur from deconvolution but is controlled with a difference of gradients loss function term. Our technique was demonstrated on a variety of patient images and on ex vivo kidney stones. Results: The noise reduction and deconvolution CNNs produced visually sharper images at low noise. In ex vivo mixed kidney stones, better visual delineation of the kidney stone components could be seen. Conclusions: A noise reduction and deconvolution CNN improves spatial resolution and reduces noise without requiring higher-resolution reference images.

Technical note: Phantom-based training framework for convolutional neural network CT noise reduction.

BACKGROUND: Deep artificial neural networks such as convolutional neural networks (CNNs) have been shown to be effective models for reducing noise in CT images while preserving anatomic details. A practical bottleneck for developing CNN-based denoising models is the procurement of training data consisting of paired examples of high-noise and low-noise CT images. Obtaining these paired data are not practical in a clinical setting where the raw projection data is not available. This work outlines a technique to optimize CNN denoising models using methods that are available in a routine clinical setting. PURPOSE: To demonstrate a phantom-based training framework for CNN noise reduction that can be efficiently implemented on any CT scanner. METHODS: The phantom-based training framework uses supervised learning in which training data are synthesized using an image-based noise insertion technique. Ten patient image series were used for training and validation (9:1) and noise-only images obtained from anthropomorphic phantom scans. Phantom noise-only images were superimposed on patient images to imitate low-dose CT images for use in training. A modified U-Net architecture was used with mean-squared-error and feature reconstruction loss. The training framework was tested for clinically indicated whole-body-low-dose CT images, as well as routine abdomen-pelvis exams for which projection data was unavailable. Performance was assessed based on root-mean-square error, structural similarity, line profiles, and visual assessment. RESULTS: When the CNN was tested on five reserved quarter-dose whole-body-low-dose CT images, noise was reduced by 75%, root-mean-square-error reduced by 34%, and structural similarity increased by 60%. Visual analysis and line profiles indicated that the method significantly reduced noise while maintaining spatial resolution of anatomic features. CONCLUSION: The proposed phantom-based training framework demonstrated strong noise reduction while preserving spatial detail. Because this method is based within the image domain, it can be easily implemented without access to projection data.

Random Search as a Neural Network Optimization Strategy for Convolutional-Neural-Network (CNN)-based Noise Reduction in CT.

In this study, we describe a systematic approach to optimize deep-learning-based image processing algorithms using random search. The optimization technique is demonstrated on a phantom-based noise reduction training framework; however, the techniques described can be applied generally for other deep learning image processing applications. The parameter space explored included number of convolutional layers, number of filters, kernel size, loss function, and network architecture (either U-Net or ResNet). A total of 100 network models were examined (50 random search, 50 ablation experiments). Following the random search, ablation experiments resulted in a very minor performance improvement indicating near optimal settings were found during the random search. The top performing network architecture was a U-Net with 4 pooling layers, 64 filters, 3×3 kernel size, ELU activation, and a weighted feature reconstruction loss (0.2×VGG + 0.8×MSE). Relative to the low-dose input image, the CNN reduced noise by 90%, reduced RMSE by 34%, and increased SSIM by 76% on six patient exams reserved for testing. The visualization of hepatic and bone lesions was greatly improved following noise reduction.

Magician's Corner: 7. Using Convolutional Neural Networks to Reduce Noise in Medical Images.

This article shows how to train a convolutional neural network to reduce noise in CT images, although the principles apply to medical and nonmedical images; authors also explore mathematical and visually weighted loss functions to adjust the appearance.

Synthesizing images from multiple kernels using a deep convolutional neural network.

PURPOSE: Filtering measured projections with a particular convolutional kernel is an essential step in analytic reconstruction of computed tomography (CT) images. A tradeoff between noise and spatial resolution exists for different choices of reconstruction kernel. In a clinical setting, this often requires producing multiple images reconstructed with different kernels for a single CT exam, which increases the burden of computation, networking, archival, and reading. We address this problem by training a deep convolutional neural network (CNN) to synthesize multiple input images into a single output image which exhibits low noise while also preserving features in images reconstructed with the sharpest kernels. METHODS: A CNN architecture consisting of repeated blocks of residual units containing a total of 20 convolutional layers was used to combine features. The CNN inputs consisted of two images produced with different reconstruction kernels, one smooth and one sharp, which were stacked in the channel dimension. The network was trained using supervised learning with both full-dose and simulated quarter-dose abdominal CT images. After training, the performance was evaluated using a reserved set of full-dose scans that were not used for network optimization. Noise reduction performance was measured by comparing root mean square (RMS) measurements in uniform regions. Spatial resolution was compared using line profiles of anatomic features. RESULTS: For the regions tested, the synthetic images feature noise levels slightly below those of the smooth input images, while maintaining the resolution of anatomic details found in the sharp input images. CONCLUSIONS: A deep CNN can be used combine features from CT images reconstructed with different kernels to produce a single synthesized image series that exhibits both low noise and high spatial resolution. This approach has implications for improving image quality, reducing radiation dose, and simplifying the clinical workflow for CT imaging.