Structural Similarity Index Measure Research Articles

GraFMRI: A graph-based fusion framework for robust multi-modal MRI reconstruction

PurposeThis study introduces GraFMRI, a novel framework designed to address the challenges of reconstructing high-quality MRI images from undersampled k-space data. Traditional methods often suffer from noise amplification and loss of structural detail, leading to suboptimal image quality. GraFMRI leverages Graph Neural Networks (GNNs) to transform multi-modal MRI data (T1, T2, PD) into a graph-based representation, enabling the model to capture intricate spatial relationships and inter-modality dependencies. MethodsThe framework integrates Graph-Based Non-Local Means (NLM) Filtering for effective noise suppression and Adversarial Training to reduce artifacts. A dynamic attention mechanism enables the model to focus on key anatomical regions, even when fully-sampled reference images are unavailable. GraFMRI was evaluated on the IXI and fastMRI datasets using Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM) as metrics for reconstruction quality. ResultsGraFMRI consistently outperforms traditional and self-supervised reconstruction techniques. Significant improvements in multi-modal fusion were observed, with better preservation of information across modalities. Noise suppression through NLM filtering and artifact reduction via adversarial training led to higher PSNR and SSIM scores across both datasets. The dynamic attention mechanism further enhanced the accuracy of the reconstructions by focusing on critical anatomical regions. ConclusionGraFMRI provides a scalable, robust solution for multi-modal MRI reconstruction, addressing noise and artifact challenges while enhancing diagnostic accuracy. Its ability to fuse information from different MRI modalities makes it adaptable to various clinical applications, improving the quality and reliability of reconstructed images.

Multi-Level Thresholding Color Image Segmentation Using Modified Gray Wolf Optimizer.

The success of image segmentation is mainly dependent on the optimal choice of thresholds. Compared to bi-level thresholding, multi-level thresholding is a more time-consuming process, so this paper utilizes the gray wolf optimizer (GWO) algorithm to address this issue and enhance accuracy. To acquire the optimal thresholds at different levels, we modify the GWO (MGWO) in terms of leader selection, position update, and mutation. We also use the Otsu method and Kapur entropy as objective functions. The performance of MGWO is compared with other color image segmentation algorithms on ten images from the BSD500 dataset in terms of objective values, variance, signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), and feature similarity index (FSIM). Experimental and non-parametric statistical analyses demonstrate that MGWO performs excellently in the multi-level thresholding segmentation of color images.

Efficient labeling for fine-tuning chest X-ray bone-suppression networks for pediatric patients.

Pneumonia, a major infectious cause of morbidity and mortality among children worldwide, is typically diagnosed using low-dose pediatric chest X-ray [CXR (chest radiography)]. In pediatric CXR images, bone occlusion leads to a risk of missed diagnosis. Deep learning-based bone-suppression networks relying on training data have enabled considerable progress to be achieved in bone suppression in adult CXR images; however, these networks have poor generalizability to pediatric CXR images because of the lack of labeled pediatric CXR images (i.e., bone images vs. soft-tissue images). Dual-energy subtraction imaging approaches are capable of producing labeled adult CXR images; however, their application is limited because they require specialized equipment, and they are infrequently employed in pediatric settings. Traditional image processing-based models can be used to label pediatric CXR images, but they are semiautomatic and have suboptimal performance. We developed an efficient labeling approach for fine-tuning pediatric CXR bone-suppression networks capable of automatically suppressing bone structures in CXR images for pediatric patients without the need for specialized equipment and technologist training. Three steps were employed to label pediatric CXR images and fine-tune pediatric bone-suppression networks: distance transform-based bone-edge detection, traditional image processing-based bone suppression, and fully automated pediatric bone suppression. In distance transform-based bone-edge detection, bone edges were automatically detected by predicting bone-edge distance-transform images, which were then used as inputs in traditional image processing. In this processing, pediatric CXR images were labeled by obtaining bone images through a series of traditional image processing techniques. Finally, the pediatric bone-suppression network was fine-tuned using the labeled pediatric CXR images. This network was initially pretrained on a public adult dataset comprising 240 adult CXR images (A240) and then fine-tuned and validated on 40 pediatric CXR images (P260_40labeled) from our customized dataset (named P260) through five-fold cross-validation; finally, the network was tested on 220 pediatric CXR images (P260_220unlabeled dataset). The distance transform-based bone-edge detection network achieved a mean boundary distance of 1.029. Moreover, the traditional image processing-based bone-suppression model obtained bone images exhibiting a relative Weber contrast of 93.0%. Finally, the fully automated pediatric bone-suppression network achieved a relative mean absolute error of 3.38%, a peak signal-to-noise ratio of 35.5dB, a structural similarity index measure of 98.1%, and a bone-suppression ratio of 90.1% on P260_40labeled. The proposed fully automated pediatric bone-suppression network, together with the proposed distance transform-based bone-edge detection network, can automatically acquire bone and soft-tissue images solely from CXR images for pediatric patients and has the potential to help diagnose pneumonia in children.

Image enhancement with art design: a visual feature approach with a CNN-transformer fusion model.

Graphic design, as a product of the burgeoning new media era, has seen its users' requirements for images continuously evolve. However, external factors such as light and noise often cause graphic design images to become distorted during acquisition. To enhance the definition of these images, this paper introduces a novel image enhancement model based on visual features. Initially, a histogram equalization (HE) algorithm is applied to enhance the graphic design images. Subsequently, image feature extraction is performed using a dual-flow network comprising convolutional neural network (CNN) and Transformer architectures. The CNN employs a residual dense block (RDB) to embed spatial local structure information with varying receptive fields. An improved attention mechanism module, attention feature fusion (AFF), is then introduced to integrate the image features extracted from the dual-flow network. Finally, through image perception quality guided adversarial learning, the model adjusts the initial enhanced image's color and recovers more details. Experimental results demonstrate that the proposed algorithm model achieves enhancement effects exceeding 90% on two large image datasets, which represents a 5%-10% improvement over other models. Furthermore, the algorithm exhibits superior performance in terms of peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) image quality evaluation metrics. Our findings indicate that the fusion model significantly enhances image quality, thereby advancing the field of graphic design and showcasing its potential in cultural and creative product design.

Film image processing and production based on high-performance calculation

In film and television production, efficient and precise image processing is vital for achieving realistic visual effects. Therefore, exploring and applying advanced image processing technologies has become an essential method for elevating the production quality of film and television projects. This work investigates the application of artificial intelligence (AI) technology in the processing and production of animated images in film and television scenarios. By comparing the performance of standard Generative Adversarial Network (GAN), DenseNet, and CycleGAN models under different noise conditions, it is found that CycleGAN performs the best in image denoising and detail restoration. Experimental results demonstrate that CycleGAN achieves a Peak Signal-to-noise Ratio (PSNR) of 30.1dB and a Structural Similarity Index Measure (SSIM) of 0.88 under Gaussian noise conditions. Moreover, CycleGAN achieves a PSNR of 29.5dB and an SSIM of 0.85 under salt-and-pepper noise conditions. It outperforms the other models in both conditions. Additionally, CycleGAN’s mean absolute error is significantly lower than that of the other models. This work demonstrates that CycleGAN can more effectively handle complex noise and generate high-quality images under unsupervised learning conditions. These findings provide new directions for future image processing research and offer important references for model selection in practical applications. This work not only offers new perspectives on the development of animation image processing technology but also establishes a theoretical foundation for applying advanced AI techniques in film and television production. Through comparative analysis of various deep learning models, this work highlights the superior performance of CycleGAN under complex noise conditions. This advancement not only drives progress in image processing technology but also provides effective solutions for efficient production and quality enhancement of future film and television works.

Deep learning based apparent diffusion coefficient map generation from multi-parametric MR images for patients with diffuse gliomas.

Apparent diffusion coefficient (ADC) maps derived from diffusion weighted magnetic resonance imaging (DWI MRI) provides functional measurements about the water molecules in tissues. However, DWI is time consuming and very susceptible to image artifacts, leading to inaccurate ADC measurements. This study aims to develop a deep learning framework to synthesize ADC maps from multi-parametric MR images. We proposed the multiparametric residual vision transformer model (MPR-ViT) that leverages the long-range context of vision transformer (ViT) layers along with the precision of convolutional operators. Residual blocks throughout the network significantly increasing the representational power of the model. The MPR-ViT model was applied to T1w and T2-fluid attenuated inversion recovery images of 501 glioma cases from a publicly available dataset including preprocessed ADC maps. Selected patients were divided into training (N=400), validation (N=50), and test (N=51) sets, respectively. Using the preprocessed ADC maps as ground truth, model performance was evaluated and compared against the Vision Convolutional Transformer (VCT) and residual vision transformer (ResViT) models with the peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), and mean squared error (MSE). The results are as follows using T1w + T2-FLAIR MRI as inputs: MPR-ViT-PSNR: 31.0±2.1, MSE: 0.009±0.0005, SSIM: 0.950±0.015. In addition, ablation studies showed the relative impact on performance of each input sequence. Both qualitative and quantitative results indicate that the proposed MR-ViT model performs favorably against the ground truth data. We show that high-quality ADC maps can be synthesized from structural MRI using a MPR-ViT model. Our predicted images show better conformality to the ground truth volume than ResViT and VCT predictions. These high-quality synthetic ADC maps would be particularly useful for disease diagnosis and intervention, especially when ADC maps have artifacts or are unavailable.

Removing atmospheric noise from InSAR interferograms in mountainous regions with a convolutional neural network

Atmospheric noise in interferometric synthetic aperture radar (InSAR)-derived estimates of surface deformation often obscures real displacement signals, especially in mountainous regions. As climate change disproportionately impacts the mountain cryosphere, a reliable technique for atmospheric correction in high-relief terrain is increasingly important. We developed and implemented a statistical machine learning atmospheric correction approach that relies on the differing spatial and topographic characteristics of slow-moving periglacial features and atmospheric noise. Our correction is applied at the native spatial and temporal resolution of the InSAR data, does not require external atmospheric reanalysis data, and can correct both stratified and turbulent atmospheric noise.Using Sentinel-1 data from 2017 to 2022, we trained a convolutional neural network (CNN) on observed atmospheric noise from 330 short-baseline interferograms and observed displacement signals from time series inversion of 1322 interferograms. We applied our trained CNN to correct 251 additional interferograms over an out-of-region application area, which were inverted to create displacement time series. We used the Rocky Mountains in New Mexico, Colorado, and Wyoming as our training, validation, testing, and application areas. When applied to our testing dataset, our correction offered performance improvements of 131%, 208%, and 68% in structural similarity index measure over corrections using atmospheric reanalysis data, phase correlation with topography, and high-pass filtering, respectively. The CNN-corrected time series reveals previously obscured kinematic behavior of rock glaciers and other features in the application dataset. Our flexible, robust approach can be used to correct arbitrary InSAR data to analyze subtle surface deformation signals for a range of science and engineering applications.

Integrating Super-Resolution with Deep Learning for Enhanced Periodontal Bone Loss Segmentation in Panoramic Radiographs.

Periodontal disease is a widespread global health concern that necessitates an accurate diagnosis for effective treatment. Traditional diagnostic methods based on panoramic radiographs are often limited by subjective evaluation and low-resolution imaging, leading to suboptimal precision. This study presents an approach that integrates Super-Resolution Generative Adversarial Networks (SRGANs) with deep learning-based segmentation models to enhance the segmentation of periodontal bone loss (PBL) areas on panoramic radiographs. By transforming low-resolution images into high-resolution versions, the proposed method reveals critical anatomical details that are essential for precise diagnostics. The effectiveness of this approach was validated using datasets from the Chungbuk National University Hospital and the Kaggle data portal, demonstrating significant improvements in both image resolution and segmentation accuracy. The SRGAN model, evaluated using the Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM) metrics, achieved a PSNR of 30.10 dB and an SSIM of 0.878, indicating high fidelity in image reconstruction. When applied to semantic segmentation using a U-Net architecture, the enhanced images resulted in a dice similarity coefficient (DSC) of 0.91 and an intersection over union (IoU) of 84.9%, compared with 0.72 DSC and 65.4% IoU for native low-resolution images. These results underscore the potential of SRGAN-enhanced imaging to improve PBL area segmentation and suggest broader applications in medical imaging, where enhanced image clarity is crucial for diagnostic accuracy. This study also highlights the importance of further research to expand the dataset diversity and incorporate clinical validation to fully realize the benefits of super-resolution techniques in medical diagnostics.

Source-detector trajectory optimization for FOV extension in dental CBCT imaging

In dental imaging, Cone Beam Computed Tomography (CBCT) is a widely used imaging modality for diagnosis and treatment planning. Small dental scanning units are the most popular due to their cost-effectiveness. However, these small systems have the limitation of a small field of view (FOV) as the source and detector move at a limited angle in a circular path. This often limits the FOV size. In this study, we addressed this issue by modifying the source-detector trajectory of the small dental device. The main goal of this study was to extend the FOV algorithmically by acquiring projection data with optimal projection angulation and isocenter location rather than upgrading any physical parts of the device. A novel algorithm to implement a Volume of Interest (VOI) guided trajectory is developed in this study based on the small dental imaging device's geometry. In addition, this algorithm is fused with a previously developed off-axis scanning method which uses an elliptical trajectory, to compensate for the existing constraints and to further extend the FOV. A comparison with standard circular trajectory is performed. The FOV of such a standard trajectory is a circle of 11 cm diameter in the axial plane. The proposed novel trajectory extends the FOV significantly and a maximum FOV of 19.5 cm is achieved with the Structural Similarity Index Measure (SSIM) score ranging between (≈98-99%) in different VOIs. The study results indicate that the proposed source-detector trajectory can extend dental imaging FOV and increase imaging performance, which ultimately results in more precise diagnosis and enhanced patient outcomes.

Deep adversarial learning models for distribution patterns of piezoelectric plate energy harvesting

This paper presents a novel approach utilizing piezoelectric plate structures with random electrode distribution patterns for energy harvesting applications across various vibration modes. For the first time, leveraging electromechanical Finite Element Analysis (eFEA) and data extraction techniques, we investigate the integration of conditional Generative Adversarial Networks (cGAN)-based dynamic models. The cGAN offers an effective technique for generating realistic synthetic data conditioned on input parameters, thereby enabling the creation of diverse and representative datasets for training energy harvesting systems. The integration of eFEA with cGAN opens up new possibilities for optimizing the design and performance of piezoelectric energy harvesters across various applications. Specifically, we explore four distinct cGAN models-based mechanics of energy harvesters by deploying distribution patterns. These models include training data generated by stacking simultaneously mode images, utilizing separate cGAN models for each mode, labeling images by mode, and concatenating all mode images into one. Our study focuses on assessing the effectiveness of these models in minimizing loss in cGAN-based power generation and predicting Structural Similarity Index Measure (SSIM) values, and more importantly, identifying the predicted data point outputs from the generated pixel image extractions. By analyzing the generated data from numerical model and its application in deep learning, we aim to enhance the understanding of the effects of distribution patterns and image processing techniques for optimal power generation and the effectiveness of piezoelectric energy harvesting systems across different vibration modes. The studies explore how different distribution patterns affect the power harvesting efficiency and frequency bandwidth, utilizing the generated datasets.

Bidirectional dynamic frame prediction network for total-body [68Ga]Ga-PSMA-11 and [68Ga]Ga-FAPI-04 PET images

PurposeTotal-body dynamic positron emission tomography (PET) imaging with total-body coverage and ultrahigh sensitivity has played an important role in accurate tracer kinetic analyses in physiology, biochemistry, and pharmacology. However, dynamic PET scans typically entail prolonged durations (\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\:\\ge\\:$$\\end{document}60 minutes), potentially causing patient discomfort and resulting in artifacts in the final images. Therefore, we propose a dynamic frame prediction method for total-body PET imaging via deep learning technology to reduce the required scanning time.MethodsOn the basis of total-body dynamic PET data acquired from 13 subjects who received [68Ga]Ga-FAPI-04 (68Ga-FAPI) and 24 subjects who received [68Ga]Ga-PSMA-11 (68Ga-PSMA), we propose a bidirectional dynamic frame prediction network that uses the initial and final 10 min of PET imaging data (frames 1–6 and frames 25–30, respectively) as inputs. The peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) were employed as evaluation metrics for an image quality assessment. Moreover, we calculated parametric images (68Ga-FAPI: \\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\:{K}_{1}$$\\end{document}, 68Ga-PSMA: \\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\:{K}_{i}$$\\end{document}) based on the supplemented sequence data to observe the quantitative accuracy of our approach. Regions of interest (ROIs) and statistical analyses were utilized to evaluate the performance of the model.ResultsBoth the visual and quantitative results illustrate the effectiveness of our approach. The generated dynamic PET images yielded PSNRs of 36.056 ± 0.709 dB for the 68Ga-PSMA group and 33.779 ± 0.760 dB for the 68Ga-FAPI group. Additionally, the SSIM reached 0.935 ± 0.006 for the 68Ga-FAPI group and 0.922 ± 0.009 for the 68Ga-PSMA group. By conducting a quantitative analysis on the parametric images, we obtained PSNRs of 36.155 ± 4.813 dB (68Ga-PSMA, \\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\:{K}_{1}$$\\end{document}) and 43.150 ± 4.102 dB (68Ga-FAPI, \\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\:{K}_{i}$$\\end{document}). The obtained SSIM values were 0.932 ± 0.041 (68Ga-PSMA) and 0.980 ± 0.011 (68Ga-FAPI). The ROI analysis conducted on our generated dynamic PET sequences also revealed that our method can accurately predict temporal voxel intensity changes, maintaining overall visual consistency with the ground truth.ConclusionIn this work, we propose a bidirectional dynamic frame prediction network for total-body 68Ga-PSMA and 68Ga-FAPI PET imaging with a reduced scan duration. Visual and quantitative analyses demonstrated that our approach performed well when it was used to predict one-hour dynamic PET images. https://github.com/OPMZZZ/BDF-NET.

Generating high-quality phase-only holograms of binary images using global loss and stochastic homogenization training strategy

Deep learning presents a highly effective for computer-generated holography (CGH). However, existing learning-based CGH algorithms may encounter significant obstacles in generating high-quality holograms for binary images. This study proposes a novel diffraction model-driven neural network, termed load sharing yielding holography (LSY-Holo), which leverages a dual-branch architecture and a global structural similarity index measure (Global_SSIM) loss function to generate phase-only holograms (POHs). LSY-Holo framework facilitates capturing global information, enhancing the network’s performance to address the ill-posed inverse problem. To further optimize the model for binary images, the research introduces a custom Binary dataset alongside a stochastic homogenization preprocessing method (SHPM), which mitigates the prevalent issue of excessive noise in reconstructed images. Additionally, this deep learning approach is integrated with an iterative algorithm, which employs holograms rapidly generated by LSY-Holo as the initial phase of the AdamW Holography (AH) algorithm. This strategy significantly accelerates the convergence of the iterative process. LSY-Holo is further extended to 3D scenes, enabling dynamic defocusing and focusing effects at varying depths. Results from numerical simulations and full-color optical reconstruction experiments demonstrate that the proposed method can enhance image quality and effectively reduce the artifacts in reconstructed images.

Counterfactual Diffusion Models for Mechanistic Explainability of Artificial Intelligence Models in Pathology.

Deep learning can extract predictive and prognostic biomarkers from histopathology whole slide images, but its interpretability remains elusive. We develop and validate MoPaDi (Morphing histoPathology Diffusion), which generates counterfactual mechanistic explanations. MoPaDi uses diffusion autoencoders to manipulate pathology image patches and flip their biomarker status by changing the morphology. Importantly, MoPaDi includes multiple instance learning for weakly supervised problems. We validate our method on four datasets classifying tissue types, cancer types within different organs, center of slide origin, and a biomarker - microsatellite instability. Counterfactual transitions were evaluated through pathologists' user studies and quantitative cell analysis. MoPaDi achieves excellent image reconstruction quality (multiscale structural similarity index measure 0.966-0.992) and good classification performance (AUCs 0.76-0.98). In a blinded user study for tissue-type counterfactuals, counterfactual images were realistic (63.3-73.3% of original images identified correctly). For other tasks, pathologists identified meaningful morphological features from counterfactual images. Overall, MoPaDi generates realistic counterfactual explanations that reveal key morphological features driving deep learning model predictions in histopathology, improving interpretability.

Adaptive deep residual network for image denoising across multiple noise levels in medical, nature, and satellite images

This research introduces the Adaptive Deep Residual Network (AdResNet), a deep convolutional neural network designed for effective image denoising in computer vision applications. Configured with the Adaptive White Shark Optimizer (AWSO), AdResNet removes noise while preserving key visual features. The model is tested on multiple noise types (Gaussian, Salt-and-Pepper, Poisson, and mixed noise) at various intensity levels, demonstrating versatility. Evaluations across medical, natural, and satellite images ensure its robustness for real-world applications. AdResNet achieves superior denoising results, with low Mean Squared Error (MSE), high Peak Signal-to-Noise Ratio (PSNR), and high Structural Similarity Index Measure (SSIM). For example, the model recorded average metrics of MSE 13.61, PSNR 48.81 dB, and SSIM 0.96 on medical images, highlighting its efficacy. These results confirm AdResNet’s suitability for applications requiring high image quality, such as medical and satellite imaging.

Investigating the Use of Traveltime and Reflection Tomography for Deep Learning-Based Sound-Speed Estimation in Ultrasound Computed Tomography.

Ultrasound computed tomography (USCT) quantifies acoustic tissue properties such as the speed-of-sound (SOS). Although full-waveform inversion (FWI) is an effective method for accurate SOS reconstruction, it can be computationally challenging for large-scale problems. Deep learning-based image-to-image learned reconstruction (IILR) methods can offer computationally efficient alternatives. This study investigates the impact of the chosen input modalities on IILR methods for high-resolution SOS reconstruction in USCT. The selected modalities are traveltime tomography (TT) and reflection tomography (RT), which produce a low-resolution SOS map and a reflectivity map, respectively. These modalities have been chosen for their lower computational cost relative to FWI and their capacity to provide complementary information: TT offers a direct SOS measure, while RT reveals tissue boundary information. Systematic analyses were facilitated by employing a virtual USCT imaging system with anatomically realistic numerical breast phantoms. Within this testbed, a supervised convolutional neural network (CNN) was trained to map dual-channel (TT and RT images) to a high-resolution SOS map. Single-input CNNs were trained separately using inputs from each modality alone (TT or RT) for comparison. The accuracy of the methods was systematically assessed using normalized root mean squared error (NRMSE), structural similarity index measure (SSIM), and peak signal-to-noise ratio (PSNR). For tumor detection performance, receiver operating characteristic analysis was employed. The dual-channel IILR method was also tested on clinical human breast data. Ensemble average of the NRMSE, SSIM, and PSNR evaluated on this clinical dataset were 0.2355, 0.8845, and 28.33 dB, respectively.

Joint image dehazing and denoising for single haze image enhancement

AbstractOutdoor haze images are typically degraded by noise due to the external environment and imaging equipment. The existing haze image enhancement methods ignore the interrelation between haze and noise, which cannot suppress the noise and remove the haze simultaneously. To address these intractable problems, a dual‐branch architecture that combines dehazing and denoising is proposed here to restore clear images. First, dark channel prior and unsupervised networks in the image dehazing branch to remove the image blur are adopted. Then, the image denoising branch removes the image noise in parallel by constructing a mean/extreme sampler and a self‐supervised network. Finally, a convolutional neural network fusion strategy is presented to fuse output images from the aforementioned two branches to generate the final qualified results. Extensive experiments reveal that the proposed haze image enhancement method outperforms other state‐of‐the‐art methods in terms of peak signal‐to‐noise ratio (PSNR) and structural similarity index measure (SSIM).

Measurement of temporal evolution of antiferromagnetic domain change in nickel oxide driven by 2TO phonon mode excitation via pump-probe experiment

The temporal evolution of antiferromagnetic domain pattern changes under resonant excitation of the second-order transverse optical phonon mode in nickel oxide was investigated using mid-infrared free electron laser (MIR-FEL) pulses and visible nanosecond probe laser pulses at room temperature. We visualized the domain patterns through magneto-optical polarized microscopy in transmission and observed their temporal variation after the phonon excitation via MIR-FEL. We evaluated the differences throughout the timeline using the structural similarity index measure technique from domain images with and without MIR-FEL irradiation. We found that the MIR-FEL irradiation induces significant structural changes in the domain patterns. The maximum difference was observed at the timing of the MIR-FEL irradiation and exponentially recovered with the time constant of 1.4 ± 0.2 ms. This rather long-time constant could be owing to the spin relaxation, whilst further investigation is needed to confirm the underlying mechanism.

Data-Driven Volumetric CT Image Generation from Surface Structures using a Patient-Specific Deep Leaning Model

PurposeOptical surface imaging presents radiation-dose-free and noninvasive approaches for image-guided radiotherapy, allowing continuous monitoring during treatment delivery. However, it falls short in cases where correlation of motion between body surface and internal tumor is complex, limiting the use of purely surface-guided surrogates for tumor tracking. Relying solely on surface-guided radiation therapy (SGRT) may not ensure accurate intra-fractional monitoring. This work aims to develop a data-driven framework, mitigating the limitations of SGRT in lung cancer radiotherapy by reconstructing volumetric CT images from surface images. Methods and MaterialsWe conducted a retrospective analysis involving 50 lung cancer patients who underwent radiotherapy and had 10-phase 4DCT scans during their treatment simulation. For each patient, we utilized nine phases of 4DCT images for patient-specific model training and validation, reserving one phase for testing purposes. Our approach employed a surface-to-volume image synthesis framework, harnessing cycle-consistency generative adversarial networks to transform surface images into volumetric representations. The framework was extensively validated using an additional 6-patient cohort with re-simulated 4DCT. ResultsThe proposed technique has produced accurate volumetric CT images from the patient's body surface. In comparison to the ground truth CT images, those generated synthetically by the proposed method exhibited the GTV center of mass difference of 1.72±0.87 mm, the overall mean absolute error of 36.2±7.0 HU, structural similarity index measure of 0.94±0.02, and Dice score coefficient of 0.81±0.07. Furthermore, the robustness of the proposed framework was found to be linked to respiratory motion. ConclusionThe proposed approach provides a novel solution to overcome the limitation of SGRT for lung cancer radiotherapy, which can potentially enable real-time volumetric imaging during radiation treatment delivery for accurate tumor tracking without radiation-induced risk. This data-driven framework offers a comprehensive solution to tackle motion management in radiotherapy, without necessitating the rigid application of first principles modeling for organ motion.

Hardware-Independent Deep Signal Processing: A Feasibility Study in Echocardiography.

Deep learning (DL) models have emerged as alternative methods to conventional ultrasound (US) signal processing, offering the potential to mimic signal processing chains, reduce inference time, and enable the portability of processing chains across hardware. This paper proposes a DL model that replicates the fine-tuned BMode signal processing chain of a high-end US system and explores the potential of using it with a different probe and a lower-end system. A deep neural network was trained in a supervised manner to map raw beamformed in-phase and quadrature component data into processed images. The dataset consisted of 30,000 cardiac image frames acquired using the GE HealthCare Vivid E95 system with the 4Vc-D matrix array probe. The signal processing chain includes depth-dependent bandpass filtering, elevation compounding, frequency compounding, and image compression and filtering. The results indicate that a lightweight DL model can accurately replicate the signal processing chain of a commercial scanner for a given application. Evaluation on a 15 patient test dataset of about three thousand image frames gave a structural similarity index measure of 98.56 ± 0.49. Applying the DL model to data from another probe showed equivalent or improved image quality. This indicates that a single DL model may be used for a set of probes on a given system that targets the same application, which could be a cost-effective tuning and implementation strategy for vendors. Further, the DL model enhanced image quality on a Verasonics dataset, suggesting the potential to port features from high-end US systems to lower-end counterparts.

Seismic Random Noise Attenuation Using DARE U-Net

Seismic data processing plays a pivotal role in extracting valuable subsurface information for various geophysical applications. However, seismic records often suffer from inherent random noise, which obscures meaningful geological features and reduces the reliability of interpretations. In recent years, deep learning methodologies have shown promising results in performing noise attenuation tasks on seismic data. In this research, we propose modifications to the standard U-Net structure by integrating dense and residual connections, which serve as the foundation of our approach named the dense and residual (DARE U-Net) network. Dense connections enhance the receptive field and ensure that information from different scales is considered during the denoising process. Our model implements local residual connections between layers within the encoder, which allows earlier layers to directly connect with deep layers. This promotes the flow of information, allowing the network to utilize filtered and unfiltered input. The combined network mechanisms preserve the spatial information loss during the contraction process so that the decoder can locate the features more accurately by retaining the high-resolution features, enabling precise location in seismic image denoising. We evaluate this adapted architecture by applying synthetic and real data sets and calculating the peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM). The effectiveness of this method is well noted.