Complex Wavelet Structural Similarity Research Articles

Time-of-flight MRA of intracranial vessels at 7 T

BackgroundThree-dimensional time-of-flight magnetic resonance angiography (TOF-MRA) is a largely adopted non-invasive technique for assessing cerebrovascular diseases. We aimed to optimize the 7-T TOF-MRA acquisition protocol, confirm that it outperforms conventional 3-T TOF-MRA, and compare 7-T TOF-MRA with digital subtraction angiography (DSA) in patients with different vascular pathologies.MethodsSeven-tesla TOF-MRA sequences with different spatial resolutions acquired in four healthy subjects were compared with 3-T TOF-MRA for signal-to-noise and contrast-to-noise ratios as well as using a qualitative scale for vessel visibility and the quantitative Canny algorithm. Four patients with cerebrovascular disease (primary arteritis of the central nervous system, saccular aneurism, arteriovenous malformation, and dural arteriovenous fistula) underwent optimized 7-T TOF-MRA and DSA as reference. Images were compared visually and using the complex-wavelet structural similarity index.ResultsContrast-to-noise ratio was higher at 7 T (4.5 ± 0.8 (mean ± standard deviation)) than at 3 T (2.7 ± 0.9). The mean quality score for all intracranial vessels was higher at 7 T (2.89) than at 3 T (2.28). Angiogram quality demonstrated a better vessel border detection at 7 T than at 3 T (44,166 versus 28,720 pixels). Of 32 parameters used for diagnosing cerebrovascular diseases on DSA, 27 (84%) were detected on 7-T TOF-MRA; the similarity index ranged from 0.52 (dural arteriovenous fistula) to 0.90 (saccular aneurysm).ConclusionsSeven-tesla TOF-MRA outperformed conventional 3-T TOF-MRA in evaluating intracranial vessels and exhibited an excellent image quality when compared to DSA. Seven-tesla TOF-MRA might improve the non-invasive diagnostic approach to several cerebrovascular diseases.Relevance statementAn optimized TOF-MRA sequence at 7 T outperforms 3-T TOF-MRA, opening perspectives to its clinical use for noninvasive diagnosis of paradigmatic pathologies of intracranial vessels.Key points• An optimized 7-T TOF-MRA protocol was selected for comparison with clinical 3-T TOF-MRA for assessing intracranial vessels.• Seven-tesla TOF-MRA outperformed 3-T TOF-MRA in both quantitative and qualitative evaluation.• Seven-tesla TOF-MRA is comparable to DSA for the diagnosis and characterization of intracranial vascular pathologies.Graphical

Shape estimation of flexible ultrasound arrays using spatial coherence: A preliminary study

A flexible ultrasound array can potentially provide a larger field-of-view, enhanced imaging resolution, and less operator dependency compared to conventional rigid transducer arrays. However, such transducer arrays require information about relative element positions for beamforming and reconstructing geometrically accurate sonograms. In this study, we assess the potential utility of using spatial coherence of backscattered radiofrequency data to estimate transducer array shape (inverse problem). The methodology is evaluated through 1) simulation of flexible arrays and 2) blinded in vivo experiments using commercial rigid transducer arrays on various anatomical targets (shoulder, forearm, scapular, posterior calf muscles, and abdomen) and multi-purpose ultrasound phantoms. The average Euclidean error of shape estimation is below 0.1 wavelengths for simulated arrays and below 1.4 wavelengths (median: 0.58 wavelengths) for real arrays. The complex wavelet structural similarity index between the B-mode images reconstructed with estimated and ground truth array shapes is above 99 % and 96 %, for simulations and experiments, respectively. These findings suggest that optimizing for spatial coherence may be an effective way to estimate the unknown shape of conformal ultrasound arrays.

ARU-GAN: U-shaped GAN based on Attention and Residual connection for super-resolution reconstruction

Plane-wave ultrasound imaging technology offers high-speed imaging but lacks image quality. To improve the image spatial resolution, beam synthesis methods are used, which often compromise the temporal resolution. Herein, we propose ARU-GAN, a super-resolution reconstruction model based on residual connectivity and attention mechanisms, to address this issue. ARU-GAN comprises a Full-scale Skip-connection U-shaped Generator (FSUG) with an attention mechanism and a Residual Attention Patch Discriminator (RAPD). The former captures global and local features of the image by using full-scale skip-connections and attention mechanisms. The latter focuses on changes in the image at different scales to enhance its discriminative ability at the patch level. ARU-GAN was trained using a combined loss function on the Plane-Wave Imaging Challenge in Medical Ultrasound (PICMUS) 2016 dataset, which includes three types of targets: point targets, cyst targets, and in-vivo targets. Compared to Coherent Plane-Wave Compounding (CPWC), ARU-GAN achieved a reduction in Full Width at Half Maximum (FWHM) by 5.78%–20.30% on point targets, improved Contrast (CR) by 7.59–11.29 percentage points, and Contrast to Noise Ratio (CNR) by 30.58%–45.22% on cyst targets. On in-vivo target, ARU-GAN improved the Peak Signal-to-Noise Ratio (PSNR) by 11.94%, the Complex-Wavelet Structural Similarity Index Measurement (CW-SSIM) by 17.11%, and the Normalized Cross Correlation (NCC) by at least 2.17% compared to existing deep learning methods. In conclusion, ARU-GAN is a promising model for the super-resolution reconstruction of plane-wave medical ultrasound images. It provides a novel solution for improving image quality, which is essential for clinical practice.

Prostatic urinary tract visualization with super-resolution deep learning models.

In urethra-sparing radiation therapy, prostatic urinary tract visualization is important in decreasing the urinary side effect. A methodology has been developed to visualize the prostatic urinary tract using post-urination magnetic resonance imaging (PU-MRI) without a urethral catheter. This study investigated whether the combination of PU-MRI and super-resolution (SR) deep learning models improves the visibility of the prostatic urinary tract. We enrolled 30 patients who had previously undergone real-time-image-gated spot scanning proton therapy by insertion of fiducial markers. PU-MRI was performed using a non-contrast high-resolution two-dimensional T2-weighted turbo spin-echo imaging sequence. Four different SR deep learning models were used: the enhanced deep SR network (EDSR), widely activated SR network (WDSR), SR generative adversarial network (SRGAN), and residual dense network (RDN). The complex wavelet structural similarity index measure (CW-SSIM) was used to quantitatively assess the performance of the proposed SR images compared to PU-MRI. Two radiation oncologists used a 1-to-5 scale to subjectively evaluate the visibility of the prostatic urinary tract. Cohen's weighted kappa (k) was used as a measure of agreement of inter-operator reliability. The mean CW-SSIM in EDSR, WDSR, SRGAN, and RDN was 99.86%, 99.89%, 99.30%, and 99.67%, respectively. The mean prostatic urinary tract visibility scores of the radiation oncologists were 3.70 and 3.53 for PU-MRI (k = 0.93), 3.67 and 2.70 for EDSR (k = 0.89), 3.70 and 2.73 for WDSR (k = 0.88), 3.67 and 2.73 for SRGAN (k = 0.88), and 4.37 and 3.73 for RDN (k = 0.93), respectively. The results suggest that SR images using RDN are similar to the original images, and the SR deep learning models subjectively improve the visibility of the prostatic urinary tract.

Pixelated volume holographic optical element for augmented reality 3D display.

Augmented reality (AR) three-dimensional (3D) display is the hardware entrance of metaverse and attracts great interest. The fusion of physical world with 3D virtual images is non-trivial. In this paper, we proposed an AR 3D display based on a pixelated volume holographic optical element (P-VHOE). The see-through combiner is prepared by spatial multiplexing. A prototype of AR 3D display with high diffraction efficiency (78.59%), high transmission (>80%) and non-repeating views is realized. Virtual 3D objects with high fidelity in depth is reconstructed by P-VHOE, with a complex wavelet structural similarity (CW-SSIM) value of 0.9882. The proposed prototype provides an efficient solution for a compact glasses-free AR 3D display. Potential applications include window display, exhibition, education, teleconference.

Affinity Propagation Based on Structural Similarity Index and Local Outlier Factor for Hyperspectral Image Clustering

In hyperspectral remote sensing, the clustering technique is an important issue of concern. Affinity propagation is a widely used clustering algorithm. However, the complex structure of the hyperspectral image (HSI) dataset presents challenge for the application of affinity propagation. In this paper, an improved version of affinity propagation based on complex wavelet structural similarity index and local outlier factor is proposed specifically for the HSI dataset. In the proposed algorithm, the complex wavelet structural similarity index is used to calculate the spatial similarity of HSI pixels. Meanwhile, the calculation strategy of the spatial similarity is simplified to reduce the computational complexity. The spatial similarity and the traditional spectral similarity of the HSI pixels jointly constitute the similarity matrix of affinity propagation. Furthermore, the local outlier factors are applied as weights to revise the original exemplar preferences of the affinity propagation. Finally, the modified similarity matrix and exemplar preferences are applied, and the clustering index is obtained by the traditional affinity propagation. Extensive experiments were conducted on three HSI datasets, and the results demonstrate that the proposed method can improve the performance of the traditional affinity propagation and provide competitive clustering results among the competitors.

Comparison of Image Enhancement Algorithms for Improving the Visual Quality in Computer Vision Application

Computer vision has its numerous real-world applications on Visual Object Tracking which includes human-computer interaction, autonomous vehicles, robotics, motion-based recognition, video indexing, surveillance and security, human-computer interaction, autonomous vehicles, robotics, motion-based recognition, video indexing, surveillance and security. The factors affecting the tracking process is due to low illumination, haze and cloudy environment and noisy environment. In this paper, we aim to extensively review the latest trends and advances in adaptive enhancement algorithm and evaluate the performance using Full reference like, SSIM (Structure Similarity Index Measure), MS-SSIM (Multi-scale Structure Similarity Index Measure), ESSIM (Edge Strength Structural Similarity Index), FSIM (Feature Similarity Index Measure), VIF (Visual Information Fidelity), CW-SSIM (complex wavelet structural similarity), UQI (Universal Quality Index), IEF (Image Enhancement Factor), IQI (Image Quality Index), EME (Enhancement Measurement Error), CVSI (Contrast and Visual Salient Information), MCSD (Multiscale contrast similarity deviation), NQM (Noise Quality Measure), Gradient Magnitude Similarity Mean (GMSM), Gradient Magnitude Similarity Deviation (GMSM) and no-reference image quality measures Perception based Image Quality Evaluator (PIQE), Blind/Reference less Image Spatial Quality Evaluator (BRISQUE), Naturalness Image Quality Evaluator (NIQE), Average Gradient (AG), Contrast, Information Entropy (IE), Lightness order Error (LOE). The main purpose of adaptive image enhancement is to smooth the uniform area and sharpen the border of an image to improve its visual quality. In this paper, fourteen image enhancement algorithms were tested on LoL dataset to benchmark the time taken to process them and their output quality was evaluated. Results from this study will give insights to image analysts for selecting image enhancement algorithms which acts as a pre- processing stage for Visual object Tracking.

Evaluation of Visualizing the Prostatic Urinary Tract in MRI With a Super Resolution Deep Learning Model for Urethra Sparing Radiotherapy

<h3>Purpose/Objective(s)</h3> In the treatment planning for urethra-sparing radiation therapy in localized prostate cancer, it is important to visualize the prostatic urinary tract to reduce the risk of urinary symptoms linked to the urethral dose. We developed a methodology for visualizing the prostatic urinary tract by post-urination magnetic resonance imaging (PU-MRI) without using a urethral catheter for urethra-sparing radiotherapy. Several super-resolution (SR) deep learning models were proposed for improving image quality. This study investigated whether these SR deep learning models improve the visibility of the prostatic urinary tract in PU-MRI. <h3>Materials/Methods</h3> We enrolled 30 patients who had previously undergone real-time-image-gated spot scanning proton therapy (RGPT) by inserting fiducial markers at our institution from October 2019 to October 2020. PU-MRI was performed using a non-contrast high-resolution two-dimensional T2-weighted turbo spin-echo (TSE) imaging sequence. Four different SR deep learning models were constructed: the EDSR (enhanced deep SR network), WDSR (widely activated SR network), SRGAN (SR generative adversarial network), and RDN (residual dense network). All models were trained using a diverse 2K resolution high-quality image dataset. We imported PU-MRIs as low-resolution images and exported four SR images that were enlarged using one of the models each. To assess the performance of the proposed SR image compared to PU-MRI, we used the complex wavelet structural similarity index measure (CW-SSIM) as the quantitative metric. A 1-to-5 scale was used to subjectively evaluate the visibility of the prostatic urinary tract by two radiation oncologists. <h3>Results</h3> The mean (± standard deviation) of the CW-SSIM in the EDSR, WDSR, SRGAN, and RDN were 0.999 ± 0.001, 0.999 ± 0.002, 0.993 ± 0.002, and 0.997 ± 0.002, respectively. The mean prostatic urinary tract visibility scores for oncologist 1 and 2 were 3.70 ± 0.64 and 3.53 ± 0.99 for PU-MRI, 3.67 ± 0.60 and 2.70 ± 0.82 for EDSR, 3.70 ± 0.53 and 2.73 ± 0.81 for WDSR, 3.67 ± 0.65 and 2.73 ± 0.81 for SRGAN, and 4.37 ± 0.61 and 3.73 ± 0.93 for RDN, respectively. <h3>Conclusion</h3> This study investigated whether the combination of PU-MRI and SR deep learning models improves visibility of the prostatic urinary tract. The results suggest that SR images using RDN are highly similar to the original image and subjectively improve the visibility of the prostatic urinary tract.

The Feasibility of Dynamic Musculoskeletal Function Analysis of the Vastus Lateralis in Endurance Runners Using Continuous, Hands-Free Ultrasound

Dynamic imaging of the skeletal muscles used to be strenuous and often impossible to perform manually. Accordingly, long-term dynamic musculoskeletal imaging has not been performed. The feasibility of long-term dynamic musculoskeletal functional analysis using hands-free ultrasound will be demonstrated in ten healthy endurance runners. After every kilometer, the vastus lateralis muscle was imaged whilst running using a fixated probe connected to a smart phone. The image quality was quantified by estimation of the probe-skin contact preservation and the field-of-view stability. Moreover, the pennation angles and muscle thicknesses were computed automatically. Long-term dynamic acquisition was successful in nine out of ten runners. Probe-skin contact loss ranged between 0 and 57% of the gait cycle. The biggest change in field-of-view occurred during the first kilometer with an average decline in complex-wavelet structural similarity index of 0.21, followed by an onward total decrease of 0.09, on average. The mean pennation angle and thickness were approximately constant, with the average fluctuation being 0.94 degrees and 0.11 cm, respectively. The feasibility of long-term musculoskeletal function analysis has been demonstrated, with probe-skin contact loss the main limiting factor. Dynamic, hands-free ultrasound might enable research for a more profound insight in the prevention and rehabilitation of musculoskeletal injuries.

Particle-Based Reconfigurable Scattering Masks for Lensless Imaging

Light scattering is typically undesired in optical systems as it often introduces defects or otherwise negatively impacts device performance. However, rather than being a hindrance, scattering can also be exploited to achieve lensless imaging using a scattering mask instead of lenses to enable devices with low-cost, compact construction, and yet a large field of view. Lensless imaging can benefit greatly from the ability to dynamically tune the scattering pattern produced by the mask; however, this often results in increased complexity and cost. Herein, we propose and demonstrate particle-based reconfigurable scattering masks to dynamically tune light scattering for lensless imaging, enabling multishot image reconstruction. Disordered particle populations are tuned by rational application of electric fields without requiring bulky or expensive components. Several assembly motifs are explored and studied for optimal performance; in particular, gold nanowires chained between planar electrodes yield the best reconstruction quality and are the main focus in this study. The distinct gold nanowire based scattering masks achieve a complex wavelet structural similarity as low as 0.36. By leveraging the submicrometer thickness of particles and the resultant large optical memory effect, an angular field of view of ±45° is demonstrated. The reconfigurable nature of the particle arrays enables multishot reconstruction which results in enhanced image quality and improved signal-to-noise ratios by up to 10-fold. These results suggest that reconfigurable particle masks could be a broadly applicable means of achieving dynamically tunable light scattering with potential applications in lensless microscopy or high-resolution imaging.

Machine Learning Diagnostic Modeling for Classifying Fibromyalgia Using B-mode Ultrasound Images.

Fibromyalgia (FM) diagnosis remains a challenge for clinicians due to a lack of objective diagnostic tools. One proposed solution is the use of quantitative ultrasound (US) techniques, such as image texture analysis, which has demonstrated discriminatory capabilities with other chronic pain conditions. From this, we propose the use of image texture variables to construct and compare two machine learning models (support vector machine [SVM] and logistic regression) for differentiating between the trapezius muscle in healthy and FM patients. US videos of the right and left trapezius muscle were acquired from healthy (n = 51) participants and those with FM (n = 57). The videos were converted into 64,800 skeletal muscle regions of interest (ROIs) using MATLAB. The ROIs were filtered by an algorithm using the complex wavelet structural similarity index (CW-SSIM), which removed ROIs that were similar. Thirty-one texture variables were extracted from the ROIs, which were then used in nested cross-validation to construct SVM and elastic net regularized logistic regression models. The generalized performance accuracy of both models was estimated and confirmed with a final validation on a holdout test set. The predicted generalized performance accuracy of the SVM and logistic regression models was computed to be 83.9 ± 2.6% and 65.8 ± 1.7%, respectively. The models achieved accuracies of 84.1%, and 66.0% on the final holdout test set, validating performance estimates. Although both machine learning models differentiate between healthy trapezius muscle and that of patients with FM, only the SVM model demonstrated clinically relevant performance levels.

Exploring the Use of Computer Vision Metrics for Spatial Pattern Comparison

Detection of changes in spatial processes has long been of interest to quantitative geographers seeking to test models, validate theories, and anticipate change. Given the current “data‐rich” environment of today, it may be time to reconsider the methodological approaches used for quantifying change in spatial processes. New tools emerging from computer vision research may hold particular potential to make significant advances in quantifying changes in spatial processes. In this article, two comparative indices from computer vision, the structural similarity (SSIM) index, and the complex wavelet structural similarity (CWSSIM) index were examined for their utility in the comparison of real and simulated spatial data sets. Gaussian Markov random fields were simulated and compared with both metrics. A case study into comparison of snow water equivalent spatial patterns over northern Canada was used to explore the properties of these indices on real‐world data. CWSSIM was found to be less sensitive than SSIM to changing window dimension. The CWSSIM appears to have significant potential in characterizing change and/or similarity; distinguishing between map pairs that possess subtle structural differences. Further research is required to explore the utility of these approaches for empirical comparison cases of different forms of landscape change and in comparison to human judgments of spatial pattern differences.

Image acquisition stability of fixated musculoskeletal sonography in an exercise setting: a quantitative analysis and comparison with freehand acquisition.

In dynamic musculoskeletal sonography, probe fixation can contribute to field of view (FOV) consistency, which is necessary for valid analysis of architectural parameters. In this volunteer study, the achieved FOV consistency in fixated ultrasonography was quantified and compared with freehand acquisition. During five resting periods during cycling exercise, longitudinal B-mode images of the vastus lateralis (VL) muscle were acquired on one thigh with a fixated probe, and by two trained observers on the other thigh. In each acquisition, the structural similarity compared to the first resting period was determined using the complex wavelet structural similarity index (CW-SSIM). Also, the pennation angle of the VL was measured. Both CW-SSIM and pennation angle were compared between fixated and freehand acquisition. Furthermore, the compression of tissue by the probe fixation was measured. In fixated acquisition, a significantly higher structural similarity (p < 0.05) and an improved repeatability of pennation angle measurement were obtained compared to freehand acquisition. Probe fixation compressed muscle tissue by 12% on average. Quantification of the structural similarity showed an increase in FOV consistency with sonography compared to freehand acquisition. The demonstrated feasibility of long-term fixated acquisition might be attractive in many medical fields and sports, and for reduction of work-related ergonomic problems among sonographers.

Making long-wave infrared face recognition robust against image quality degradations

ABSTRACTFace identification systems that operate on long wave infrared (LWIR) images are able to overcome some of the limitations of approaches based on visible images, such as dealing effectively with illumination variations. Nonetheless, distortions of perceptual image quality can impair the performance of thermal face recognition systems. Although the interaction between perceptual image quality and tasks such as face detection has been studied on visual images, the development of similar models has not been applied to the LWIR-based face recognition problem. Here, we analyze the impact of four common infrared image distortions (gaussian noise, blur, non-uniformity, and JPEG compression) on two thermal face recognition systems. We propose an LWIR image face recognition framework, based on thermal signature templates, and enhanced by natural scene statistics image quality descriptors, which achieves system robustness against image quality distortions. Furthermore, we develop a novel infrared face recognition system that is based on the complex wavelet structural similarity (CW-SSIM) index, which exhibits resistance to image distortions, within a relatively simple implementation. Our results validate the applicability of image quality assessment models to biometric tasks on LWIR images. To facilitate our study, we created two new LWIR facial image databases, with different poses, expressions and illumination conditions.

An Improved PSO-Based Multilevel Image Segmentation Technique Using Minimum Cross-Entropy Thresholding

Entropy-based thresholding techniques are quite popular and effective for image segmentation. Among different entropy-based techniques, minimum cross-entropy thresholding (MCET) has received wide attention in the field of image segmentation. Considering the high time complexity of MCET technique for multilevel thresholding, recursive approach to reducing its computational cost is highly desired. To reduce the complexity, further optimization techniques can be applied to find optimal multilevel threshold values. In this paper, a novel improved particle swarm optimization (IPSO)-based multilevel thresholding algorithm is proposed to search the near-optimal MCET thresholds. The general PSO algorithm often suffers from premature convergence problem which has been addressed in the IPSO by decomposing a high-dimensional swarm into several one-dimensional swarms, and then premature convergence is removed from each one-dimensional swarm. The proposed technique is applied to the set of grayscale images, and the experimental results infer that it produces better MCET optimal threshold values at a higher and faster convergence rate. The qualitative and quantitative results are compared with existing optimization techniques like modified artificial bee colony, Cuckoo search, Firefly, particle swarm optimization, and genetic algorithm. It has been observed that the proposed technique performs better in terms of producing better fitness value, less CPU time as quantitative measurements, and effective misclassification error, peak signal-to-noise ratio, feature similarity index measurement, complex wavelet structural similarity index measurement values as qualitative measurements compared to other considered state-of-the-art methods.

Similarity measurement based rest position re-initialization in the MRI vocal tract image sequences

Magnetic resonance imaging is often used for the speech production research. One important aspect is to segment the vocal tract in the image sequence. However, during the deformation of the vocal tract, it is highly possible that the landmark deviates from its correct position and is in need of reinitialization. During the natural speech production of the subject, the vocal tract may return to its initial position during the pause at the end of a sentence. This could be used to reset the landmark to its correct position. In order to determine the pause in the image sequence, we used similarity based measurements to compare the similarity between the current frame and the first frame, which is the beginning of a sentence and the vocal tract is at the rest position. These measurements include Structural Similarity (SSIM), Complex Wavelet Structural Similarity (CW-SSIM), Visual Information Fidelity in Pixel (VIFP), Peak Noise to Signal Ratio (PSNR), etc. We found CW-SSIM outperformed the other methods. We calculated the similarity measurements and they varied periodically during the speech. CW-SSIM returned to a maximum of around 0.9, which indicated that the vocal tract returned to its initial position. The rest of the similarity measurements returned to a maximum value greatly deviated from 1, which indicated that the CW-SSIM was the best candidate.

Learning quality assessment of retargeted images

Content-aware image resizing (or image retargeting) enables images to be fit to different display devices having different aspect ratios while preserving salient image content. There are many approaches to retargeting, although no “best” method has been agreed upon. Therefore, finding ways to assess the quality of image retargeting has become a prominent challenge. Traditional image quality assessment methods are not directly applicable to image retargeting because the retargeted image size is not same as the original one. In this paper, we propose an open framework for image retargeting quality assessment, where the quality prediction engine is a trained Radial Basis Function (RBF) neural network. Broadly, our approach is motivated by the observation that no single method can be expected to perform well on all types of content. We train the network on ten perceptually relevant features, including a saliency-weighted, SIFT-directed complex wavelet structural similarity (CW-SSIM) index, and a new image aesthetics evaluation method. These two features and eight other features are used by the neural network to learn to assess the quality of retargeted images. The accuracy of the new model is extensively verified by simulations.

SU‐G‐JeP1‐03: Automatic Motion Tracking Reset in Ultrasound Liver Image Sequences

Purpose:We explore an automatic anatomical landmarks tracking reset method in the ultrasound liver sequences, with the goal to improve the robustness and accuracy of the tracking.Methods:The proposed tracking failure re‐initialization method works as follows: before anatomic landmarks tracking is carried out, the image similarity coefficient is calculated (In our method, the complex wavelet structural similarity is used), between the region of interest (ROI) selected in current frame and the ROI in the first frame. If this similarity coefficient exceeds a set threshold, the positions of anatomical landmarks are reset to those which were input for the first frame. This provides a method to prevent accumulation of errors over long sequences, which can lead to erroneous tracking, and amounts to a sort of “automatic reset” of the tracking starting points based on initial a priori information.Results:The experiments are conducted to track the landmarks in the 5 volunteer B‐mode ultrasound liver sequences. The results were evaluated by comparison with manual annotations of liver feature (vessels) throughout each image sequence which were provided after automated tracking was complete. Tracking accuracy was evaluated using the Euclidean distance between tracked points and manually annotated points which was summarized by the mean. The mean error for all ROIs in different sequences are about 1–1.5 mm with the proposed automatic reset method, while the error is about 2 mm without the automatic reset method.Conclusion:Due to the periodic motion of the anatomical landmarks, the proposed image similarity‐based automatic re‐initialization method can improve the accuracy and robustness of the motion tracking in the ultrasound liver sequences. It may also be applied to other organs’ tracking problem using ultrasound imaging.

Multimodal gray image fusion metric based on complex wavelet structural similarity

A novel multimodal gray image fusion metric is proposed based on complex wavelet structural similarity (CW-SSIM). The idea behind is that consistent phase changes with the wavelet coefficients corresponding to certain distortions in the image. Comparing with other common objective fusion metrics, the proposed metric correspond better to the subjective evaluation in the experiments.

C-DIIVINE: No-reference image quality assessment based on local magnitude and phase statistics of natural scenes

It is widely known that the wavelet coefficients of natural scenes possess certain statistical regularities which can be affected by the presence of distortions. The DIIVINE (Distortion Identification-based Image Verity and Integrity Evaluation) algorithm is a successful no-reference image quality assessment (NR IQA) algorithm, which estimates quality based on changes in these regularities. However, DIIVINE operates based on real-valued wavelet coefficients, whereas the visual appearance of an image can be strongly determined by both the magnitude and phase information.In this paper, we present a complex extension of the DIIVINE algorithm (called C-DIIVINE), which blindly assesses image quality based on the complex Gaussian scale mixture model corresponding to the complex version of the steerable pyramid wavelet transform. Specifically, we applied three commonly used distribution models to fit the statistics of the wavelet coefficients: (1) the complex generalized Gaussian distribution is used to model the wavelet coefficient magnitudes, (2) the generalized Gaussian distribution is used to model the coefficients׳ relative magnitudes, and (3) the wrapped Cauchy distribution is used to model the coefficients׳ relative phases. All these distributions have characteristic shapes that are consistent across different natural images but change significantly in the presence of distortions. We also employ the complex wavelet structural similarity index to measure degradation of the correlations across image scales, which serves as an important indicator of the subbands׳ energy distribution and the loss of alignment of local spectral components contributing to image structure. Experimental results show that these complex extensions allow C-DIIVINE to yield a substantial improvement in predictive performance as compared to its predecessor, and highly competitive performance relative to other recent no-reference algorithms.