Interpolation-based image super-resolution using multisurface fitting

Enhancing Knee MR Image Clarity through Image Domain Super-Resolution Reconstruction

This study introduces a hybrid analytical super-resolution (SR) pipeline aimed at enhancing the resolution of medical magnetic resonance imaging (MRI) scans. The primary objective is to overcome the limitations of clinical MRI resolution without the need for additional expensive hardware. The proposed pipeline involves three key steps: pre-processing to re-slice and register the image stacks; SR reconstruction to combine information from three orthogonal image stacks to generate a high-resolution image stack; and post-processing using an artefact reduction convolutional neural network (ARCNN) to reduce the block artefacts introduced during SR reconstruction. The workflow was validated on a dataset of six knee MRIs obtained at high resolution using various sequences. Quantitative analysis of the method revealed promising results, showing an average mean error of 1.40 ± 2.22% in voxel intensities between the SR denoised images and the original high-resolution images. Qualitatively, the method improved out-of-plane resolution while preserving in-plane image quality. The hybrid SR pipeline also displayed robustness across different MRI sequences, demonstrating potential for clinical application in orthopaedics and beyond. Although computationally intensive, this method offers a viable alternative to costly hardware upgrades and holds promise for improving diagnostic accuracy and generating more anatomically accurate models of the human body.

Medical applications of generative adversarial network: a visualization analysis

BACKGROUND

Deep learning (DL) is one of the latest approaches to artificial intelligence. As an unsupervised DL method, a generative adversarial network (GAN) can be used to synthesize new data.

PURPOSE

To explore GAN applications in medicine and point out the significance of its existence for clinical medical research, as well as to provide a visual bibliometric analysis of GAN applications in the medical field in combination with the scientometric software Citespace and statistical analysis methods.

MATERIAL AND METHODS

PubMed, MEDLINE, Web of Science, and Google Scholar were searched to identify studies of GAN in medical applications between 2017 and 2022. This study was performed and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Citespace was used to analyze the number of publications, authors, institutions, and keywords of articles related to GAN in medical applications.

RESULTS

The applications of GAN in medicine are not limited to medical image processing, but will also penetrate wider and more complex fields, or may be applied to clinical medicine. Eligibility criteria were the full texts of peer-reviewed journals reporting the application of GANs in medicine. Research selections included material published in English between 1 January 2017 and 1 December 2022.

CONCLUSION

GAN has been fully applied to the medical field and will be more deeply and widely used in clinical medicine, especially in the field of privacy protection and medical diagnosis. However, clinical applications of GAN require consideration of ethical and legal issues. GAN-based applications should be well validated by expert radiologists.

RankSRGAN: Super Resolution Generative Adversarial Networks With Learning to Rank

Generative Adversarial Networks (GAN) have demonstrated the potential to recover realistic details for single image super-resolution (SISR). To further improve the visual quality of super-resolved results, PIRM2018-SR Challenge employed perceptual metrics to assess the perceptual quality, such as PI, NIQE, and Ma. However, existing methods cannot directly optimize these indifferentiable perceptual metrics, which are shown to be highly correlated with human ratings. To address the problem, we propose Super-Resolution Generative Adversarial Networks with Ranker (RankSRGAN) to optimize generator in the direction of different perceptual metrics. Specifically, we first train a Ranker which can learn the behaviour of perceptual metrics and then introduce a novel rank-content loss to optimize the perceptual quality. The most appealing part is that the proposed method can combine the strengths of different SR methods to generate better results. Furthermore, we extend our method to multiple Rankers to provide multi-dimension constraints for the generator. Extensive experiments show that RankSRGAN achieves visually pleasing results and reaches state-of-the-art performance in perceptual metrics and quality. Project page: https://wenlongzhang0517.github.io/Projects/RankSRGAN.

Super-Resolution Reconstruction of Cytoskeleton Image Based on A-Net Deep Learning Network

To date, live-cell imaging at the nanometer scale remains challenging. Even though super-resolution microscopy methods have enabled visualization of sub-cellular structures below the optical resolution limit, the spatial resolution is still far from enough for the structural reconstruction of biomolecules in vivo (i.e., ~24 nm thickness of microtubule fiber). In this study, a deep learning network named A-net was developed and shows that the resolution of cytoskeleton images captured by a confocal microscope can be significantly improved by combining the A-net deep learning network with the DWDC algorithm based on a degradation model. Utilizing the DWDC algorithm to construct new datasets and taking advantage of A-net neural network's features (i.e., considerably fewer layers and relatively small dataset), the noise and flocculent structures which originally interfere with the cellular structure in the raw image are significantly removed, with the spatial resolution improved by a factor of 10. The investigation shows a universal approach for exacting structural details of biomolecules, cells and organs from low-resolution images.

Deep Learning-Based Super-Resolution Reconstruction and Algorithm Acceleration of Mars Hyperspectral CRISM Data

In Mars exploration, hyper-spectrometry plays an important role due to its high spectral resolution. However, due to the technical difficulty and the data size, the spatial resolution or the coverage of hyperspectral data is often limited. This limitation can be alleviated by deep learning-based super-resolution (SR) reconstruction. But the spatial size and batch size of the input training data is limited due to the large number of spectral channels. To improve the efficiency of model training and SR reconstruction, a dataset based on CRISM hyperspectral data is created in this paper, and its redundancy is analyzed in both spectral and spatial spital dimensions. Compression algorithms based on data selection and PCA are used to reduce the size of the input training data. A network that can perform spatial SR and spectral enhancement is also proposed to make the network can be trained with the compressed data. With these compression algorithms and network, high-resolution data with 235 bands can be reconstructed from the low-resolution data with only 40 bands. Compared with the network trained on the original low-resolution data with 235 bands, the model training time and the SR reconstruction runtime can be reduced to 30% and 23% with practically no accuracy loss. The effectiveness of compression algorithms based on data selection also indicates that maybe not all the bands need to be transmitted from the Mars probes or be collected. Furthermore, it would, in principle, help improve the efficiency of satellite data transmission and simplify the design of the hyper-spectrometer. Additionally, a method for spatial dimension correlation evaluation is also proposed in this paper. The spatial compression shows that the proposed method can reflect the correlation of spatial texture between patches, and the model can be acceptably trained with only half of the original data.

Learning Local Distribution for Extremely Efficient Single-Image Super-Resolution

Achieving balance between efficiency and performance is a key problem for convolution neural network (CNN)-based single-image super-resolution (SISR) algorithms. Existing methods tend to directly output high-resolution (HR) pixels or residuals to reconstruct the HR image and focus a lot of attention on designing powerful CNN backbones. However, this reconstruction way requires the CNN backbone to have good ability to fit the mapping function from LR pixels to HR pixels, which certainly held these methods back from achieving extreme efficiency and from working in embedded environments. In this work, we propose a novel distribution learning architecture to estimate the local distribution and reconstruct HR pixels by sampling the local distribution with the corresponding 2D coordinates. We also improve the backbone structure to better support the proposed distribution learning architecture. The experimental results demonstrate that the proposed method achieves state-of-the-art performance for extremely efficient SISR and exhibits a good balance between efficiency and performance.

Super-Resolution Ultrasound Imaging Scheme Based on a Symmetric Series Convolutional Neural Network

In this paper, we propose a symmetric series convolutional neural network (SS-CNN), which is a novel deep convolutional neural network (DCNN)-based super-resolution (SR) technique for ultrasound medical imaging. The proposed model comprises two parts: a feature extraction network (FEN) and an up-sampling layer. In the FEN, the low-resolution (LR) counterpart of the ultrasound image passes through a symmetric series of two different DCNNs. The low-level feature maps obtained from the subsequent layers of both DCNNs are concatenated in a feed forward manner, aiding in robust feature extraction to ensure high reconstruction quality. Subsequently, the final concatenated features serve as an input map to the latter 2D convolutional layers, where the textural information of the input image is connected via skip connections. The second part of the proposed model is a sub-pixel convolutional (SPC) layer, which up-samples the output of the FEN by multiplying it with a multi-dimensional kernel followed by a periodic shuffling operation to reconstruct a high-quality SR ultrasound image. We validate the performance of the SS-CNN with publicly available ultrasound image datasets. Experimental results show that the proposed model achieves a high-quality reconstruction of the ultrasound image over the conventional methods in terms of peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM), while providing compelling SR reconstruction time.

TDPN: Texture and Detail-Preserving Network for Single Image Super-Resolution

Single image super-resolution (SISR) using deep convolutional neural networks (CNNs) achieves the state-of-the-art performance. Most existing SISR models mainly focus on pursuing high peak signal-to-noise ratio (PSNR) and neglect textures and details. As a result, the recovered images are often perceptually unpleasant. To address this issue, in this paper, we propose a texture and detail-preserving network (TDPN), which focuses not only on local region feature recovery but also on preserving textures and details. Specifically, the high-resolution image is recovered from its corresponding low-resolution input in two branches. First, a multi-reception field based branch is designed to let the network fully learn local region features by adaptively selecting local region features in different reception fields. Then, a texture and detail-learning branch supervised by the textures and details decomposed from the ground-truth high resolution image is proposed to provide additional textures and details for the super-resolution process to improve the perceptual quality. Finally, we introduce a gradient loss into the SISR field and define a novel hybrid loss to strengthen boundary information recovery and to avoid overly smooth boundary in the final recovered high-resolution image caused by using only the MAE loss. More importantly, the proposed method is model-agnostic, which can be applied to most off-the-shelf SISR networks. The experimental results on public datasets demonstrate the superiority of our TDPN on most state-of-the-art SISR methods in PSNR, SSIM and perceptual quality. We will share our code on https://github.com/tocaiqing/TDPN.

Multi-Path Deep CNN with Residual Inception Network for Single Image Super-Resolution

Recent research on single-image super-resolution (SISR) using deep convolutional neural networks has made a breakthrough and achieved tremendous performance. Despite their significant progress, numerous convolutional neural networks (CNN) are limited in practical applications, owing to the requirement of the heavy computational cost of the model. This paper proposes a multi-path network for SISR, known as multi-path deep CNN with residual inception network for single image super-resolution. In detail, a residual/ResNet block with an Inception block supports the main framework of the entire network architecture. In addition, remove the batch normalization layer from the residual network (ResNet) block and max-pooling layer from the Inception block to further reduce the number of parameters to preventing the over-fitting problem during the training. Moreover, a conventional rectified linear unit (ReLU) is replaced with Leaky ReLU activation function to speed up the training process. Specifically, we propose a novel two upscale module, which adopts three paths to upscale the features by jointly using deconvolution and upsampling layers, instead of using single deconvolution layer or upsampling layer alone. The extensive experimental results on image super-resolution (SR) using five publicly available test datasets, which show that the proposed model not only attains the higher score of peak signal-to-noise ratio/structural similarity index matrix (PSNR/SSIM) but also enables faster and more efficient calculations against the existing image SR methods. For instance, we improved our method in terms of overall PSNR on the SET5 dataset with challenging upscale factor 8× as 1.88 dB over the baseline bicubic method and reduced computational cost in terms of number of parameters 62% by deeply-recursive convolutional neural network (DRCN) method.

License Plate Image Reconstruction Based on Generative Adversarial Networks

License plate image reconstruction plays an important role in Intelligent Transportation Systems. In this paper, a super-resolution image reconstruction method based on Generative Adversarial Networks (GAN) is proposed. The proposed method mainly consists of four parts: (1) pretreatment for the input image; (2) image features extraction using residual dense network; (3) introduction of progressive sampling, which can provide larger receptive field and more information details; (4) discriminator based on markovian discriminator (PatchGAN) can make a more accurate judgment, which guides the generator to reconstruct images with higher quality and details. Regarding the Chinese City Parking Dataset (CCPD) dataset, compared with the current better algorithm, the experiment results prove that our model has a higher peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) and less reconstruction time, which verifies the feasibility of our approach.

Multimodal Deep Unfolding for Guided Image Super-Resolution

The reconstruction of a high resolution image given a low resolution observation is an ill-posed inverse problem in imaging. Deep learning methods rely on training data to learn an end-to-end mapping from a low-resolution input to a highresolution output. Unlike existing deep multimodal models that do not incorporate domain knowledge about the problem, we propose a multimodal deep learning design that incorporates sparse priors and allows the effective integration of information from another image modality into the network architecture. Our solution relies on a novel deep unfolding operator, performing steps similar to an iterative algorithm for convolutional sparse coding with side information; therefore, the proposed neural network is interpretable by design. The deep unfolding architecture is used as a core component of a multimodal framework for guided image super-resolution. An alternative multimodal design is investigated by employing residual learning to improve the training efficiency. The presented multimodal approach is applied to super-resolution of near-infrared and multi-spectral images as well as depth upsampling using RGB images as side information. Experimental results show that our model outperforms state-ofthe-art methods.

Super-Resolution Reconstruction Algorithm for Infrared Image with Double Regular Items Based on Sub-Pixel Convolution

In this paper, an adaptive dual-regularization super-resolution reconstruction algorithm based on sub-pixel convolution (MPSR) is proposed. There are two novel features of the algorithm: First, the traditional regularization algorithm and sub-pixel convolution algorithm are combined to enrich the details; then, a regularization function with two adaptive parameters and two regularization terms is proposed to enhance the edge. MPSR firstly enhances the multi-scale detail of low-resolution images; then, regular processing and feature extraction are carried out; finally, sub-pixel convolution is used to fuse the extracted features to generate high-resolution images. The experimental results show that the subjective and objective evaluation indexes (PSNR/SSIM) of the algorithm have achieved satisfactory results.

Single Image Super-Resolution: Depthwise Separable Convolution Super-Resolution Generative Adversarial Network

The super-resolution generative adversarial network (SRGAN) is a seminal work that is capable of generating realistic textures during single image super-resolution. However, the hallucinated details are often accompanied by unpleasant artifacts. To further enhance the visual quality, we propose a deep learning method for single image super-resolution (SR). Our method directly learns an end-to-end mapping between the low/high-resolution images. The method is based on depthwise separable convolution super-resolution generative adversarial network (DSCSRGAN). A new depthwise separable convolution dense block (DSC Dense Block) was designed for the generator network, which improved the ability to represent and extract image features, while greatly reducing the total amount of parameters. For the discriminator network, the batch normalization (BN) layer was discarded, and the problem of artifacts was reduced. A frequency energy similarity loss function was designed to constrain the generator network to generate better super-resolution images. Experiments on several different datasets showed that the peak signal-to-noise ratio (PSNR) was improved by more than 3 dB, structural similarity index (SSIM) was increased by 16%, and the total parameter was reduced to 42.8% compared with the original model. Combining various objective indicators and subjective visual evaluation, the algorithm was shown to generate richer image details, clearer texture, and lower complexity.

A Comparable Study of CNN-Based Single Image Super-Resolution for Space-Based Imaging Sensors

In the case of space-based space surveillance (SBSS), images of the target space objects captured by space-based imaging sensors usually suffer from low spatial resolution due to the extremely long distance between the target and the imaging sensor. Image super-resolution is an effective data processing operation to get informative high resolution images. In this paper, we comparably study four recent popular models for single image super-resolution based on convolutional neural networks (CNNs) with the purpose of space applications. We specially fine-tune the super-resolution models designed for natural images using simulated images of space objects, and test the performance of different CNN-based models in different conditions that are mainly considered for SBSS. Experimental results show the advantages and drawbacks of these models, which could be helpful for the choice of proper CNN-based super-resolution method to deal with image data of space objects.

Infrared Image Super-Resolution Reconstruction Based on Quaternion Fractional Order Total Variation with Lp Quasinorm

Owing to the limitations of the imaging principle as well as the properties of imaging systems, infrared images often have some drawbacks, including low resolution, a lack of detail, and indistinct edges. Therefore, it is essential to improve infrared image quality. Considering the information of neighbors, a description of sparse edges, and by avoiding staircase artifacts, a new super-resolution reconstruction (SRR) method is proposed for infrared images, which is based on fractional order total variation (FTV) with quaternion total variation and the L p quasinorm. Our proposed method improves the sparsity exploitation of FTV, and efficiently preserves image structures. Furthermore, we adopt the plug-and-play alternating direction method of multipliers (ADMM) and the fast Fourier transform (FFT) theory for the proposed method to improve the efficiency and robustness of our algorithm; in addition, an accelerated step is adopted. Our experimental results show that the proposed method leads to excellent performances in terms of an objective evaluation and the subjective visual effect.

Super-resolution CT Image Reconstruction Based on Dictionary Learning and Sparse Representation

In this paper, a single-computed tomography (CT) image super-resolution (SR) reconstruction scheme is proposed. This SR reconstruction scheme is based on sparse representation theory and dictionary learning of low- and high-resolution image patch pairs to improve the poor quality of low-resolution CT images obtained in clinical practice using low-dose CT technology. The proposed strategy is based on the idea that image patches can be well represented by sparse coding of elements from an overcomplete dictionary. To obtain similarity of the sparse representations, two dictionaries of low- and high-resolution image patches are jointly trained. Then, sparse representation coefficients extracted from the low-resolution input patches are used to reconstruct the high-resolution output. Sparse representation is used such that the trained dictionary pair can reduce computational costs. Combined with several appropriate iteration operations, the reconstructed high-resolution image can attain better image quality. The effectiveness of the proposed method is demonstrated using both clinical CT data and simulation image data. Image quality evaluation indexes (root mean squared error (RMSE) and peak signal-to-noise ratio (PSNR)) indicate that the proposed method can effectively improve the resolution of a single CT image.

Color Fringe Correction by the Color Difference Prediction Using the Logistic Function

This paper proposes a new color fringe correction method that preserves the object color well by the color difference prediction using the logistic function. We observe two characteristics between normal edge (NE) and degraded edge (DE) due to color fringe: 1) the DE has relatively smaller R-G and B-G correlations than the NE and 2) the color difference in the NE can be fitted by the logistic function. The proposed method adjusts the color difference of the DE to the logistic function by maximizing the R-G and B-G correlations in the corrected color fringe image. The generalized logistic function with four parameters requires a high computational load to select the optimal parameters. In experiments, a one-parameter optimization can correct color fringe gracefully with a reduced computational load. Experimental results show that the proposed method restores well the original object color in the DE, whereas existing methods give monochromatic or distorted color.

Second-Order Regression-Based MR Image Upsampling

The spatial resolution of magnetic resonance imaging (MRI) is often limited due to several reasons, including a short data acquisition time. Several advanced interpolation-based image upsampling algorithms have been developed to increase the resolution of MR images. These methods estimate the voxel intensity in a high-resolution (HR) image by a weighted combination of voxels in the original low-resolution (LR) MR image. As these methods fall into the zero-order point estimation framework, they only include a local constant approximation of the image voxel and hence cannot fully represent the underlying image structure(s). To this end, we extend the existing zero-order point estimation to higher orders of regression, allowing us to approximate a mapping function between local LR-HR image patches by a polynomial function. Extensive experiments on open-access MR image datasets and actual clinical MR images demonstrate that our algorithm can maintain sharp edges and preserve fine details, while the current state-of-the-art algorithms remain prone to some visual artifacts such as blurring and staircasing artifacts.

Sparse Representation-Based Multiple Frame Video Super-Resolution

In this paper, we propose two multiple-frame super-resolution (SR) algorithms based on dictionary learning (DL) and motion estimation. First, we adopt the use of video bilevel DL, which has been used for single-frame SR. It is extended to multiple frames by using motion estimation with sub-pixel accuracy. We propose a batch and a temporally recursive multi-frame SR algorithm, which improves over single-frame SR. Finally, we propose a novel DL algorithm utilizing consecutive video frames, rather than still images or individual video frames, which further improves the performance of the video SR algorithms. Extensive experimental comparisons with the state-of-the-art SR algorithms verify the effectiveness of our proposed multiple-frame video SR approach.

Single image superresolution based on gradient profile sharpness

Single image superresolution is a classic and active image processing problem, which aims to generate a high-resolution (HR) image from a low-resolution input image. Due to the severely under-determined nature of this problem, an effective image prior is necessary to make the problem solvable, and to improve the quality of generated images. In this paper, a novel image superresolution algorithm is proposed based on gradient profile sharpness (GPS). GPS is an edge sharpness metric, which is extracted from two gradient description models, i.e., a triangle model and a Gaussian mixture model for the description of different kinds of gradient profiles. Then, the transformation relationship of GPSs in different image resolutions is studied statistically, and the parameter of the relationship is estimated automatically. Based on the estimated GPS transformation relationship, two gradient profile transformation models are proposed for two profile description models, which can keep profile shape and profile gradient magnitude sum consistent during profile transformation. Finally, the target gradient field of HR image is generated from the transformed gradient profiles, which is added as the image prior in HR image reconstruction model. Extensive experiments are conducted to evaluate the proposed algorithm in subjective visual effect, objective quality, and computation time. The experimental results demonstrate that the proposed approach can generate superior HR images with better visual quality, lower reconstruction error, and acceptable computation efficiency as compared with state-of-the-art works.

Image quality assessment based on inter-patch and intra-patch similarity

In this paper, we propose a full-reference (FR) image quality assessment (IQA) scheme, which evaluates image fidelity from two aspects: the inter-patch similarity and the intra-patch similarity. The scheme is performed in a patch-wise fashion so that a quality map can be obtained. On one hand, we investigate the disparity between one image patch and its adjacent ones. This disparity is visually described by an inter-patch feature, where the hybrid effect of luminance masking and contrast masking is taken into account. The inter-patch similarity is further measured by modifying the normalized correlation coefficient (NCC). On the other hand, we also attach importance to the impact of image contents within one patch on the IQA problem. For the intra-patch feature, we consider image curvature as an important complement of image gradient. According to local image contents, the intra-patch similarity is measured by adaptively comparing image curvature and gradient. Besides, a nonlinear integration of the inter-patch and intra-patch similarity is presented to obtain an overall score of image quality. The experiments conducted on six publicly available image databases show that our scheme achieves better performance in comparison with several state-of-the-art schemes.