Search

Multispectral satellite imaging sensors acquire various spectral band images and have a unique spectroscopic property in each band. Unfortunately, image artifacts from imaging sensor noise often affect the quality of scenes and have a negative impact on applications for satellite imagery. Recently, deep learning approaches have been extensively explored to remove noise in satellite imagery. Most deep learning denoising methods, however, follow a supervised learning scheme, which requires matched noisy image and clean image pairs that are difficult to collect in real situations. In this article, we propose a novel unsupervised multispectral denoising method for satellite imagery using a wavelet directional cycle-consistent adversarial network (WavCycleGAN). The proposed method is based on an unsupervised learning scheme using adversarial loss and cycle-consistency loss to overcome the lack of paired data. Moreover, in contrast to the standard image-domain cycleGAN, we introduce a wavelet directional learning scheme for effective denoising without sacrificing high-frequency components such as edges and detailed information. Experimental results for the removal of vertical stripes and wave noise in satellite imaging sensors demonstrate that the proposed method effectively removes noise and preserves important high-frequency features of satellite images.

(1) To compare low-contrast detectability of a deep learning–based denoising algorithm (DLA) with ADMIRE and FBP, and (2) to compare image quality parameters of DLA with those of reconstruction methods from two different CT vendors (ADMIRE, IMR, and FBP). Using abdominal CT images of 100 patients reconstructed via ADMIRE and FBP, we trained DLA by feeding FBP images as input and ADMIRE images as the ground truth. To measure the low-contrast detectability, the randomized repeat scans of Catphan® phantom were performed under various conditions of radiation exposures. Twelve radiologists evaluated the presence/absence of a target on a five-point confidence scale. The multi-reader multi-case area under the receiver operating characteristic curve (AUC) was calculated, and non-inferiority tests were performed. Using American College of Radiology CT accreditation phantom, contrast-to-noise ratio, target transfer function, noise magnitude, and detectability index (d’) of DLA, ADMIRE, IMR, and FBPs were computed. The AUC of DLA in low-contrast detectability was non-inferior to that of ADMIRE (p < .001) and superior to that of FBP (p < .001). DLA improved the image quality in terms of all physical measurements compared to FBPs from both CT vendors and showed profiles of physical measurements similar to those of ADMIRE. The low-contrast detectability of the proposed deep learning–based denoising algorithm was non-inferior to that of ADMIRE and superior to that of FBP. The DLA could successfully improve image quality compared with FBP while showing the similar physical profiles of ADMIRE. • Low-contrast detectability in the images denoised using the deep learning algorithm was non-inferior to that in the images reconstructed using standard algorithms. • The proposed deep learning algorithm showed similar profiles of physical measurements to advanced iterative reconstruction algorithm (ADMIRE).

Deep-learning-based tomographic imaging is an important application of artificial intelligence and a new frontier of machine learning. Deep learning has been widely used in computer vision and image analysis, which deal with existing images, improve these images, and produce features from them. Since 2016, deep learning techniques have been actively researched for tomographic imaging, especially in the context of biomedicine, with impressive results and great potential. Tomographic reconstruction produces images of multi-dimensional structures from externally measured ‘encoded’ data in the form of various tomographic transforms (integrals, harmonics, echoes and so on). In this Review, we provide a general background, highlight representative results with an emphasis on medical imaging, and discuss key issues that need to be addressed in this emerging field. In particular, tomographic imaging is an integral part of modern medicine, and will play a key role in personalized, preventive and precision medicine and make it intelligent, inexpensive and indiscriminate. The popularity of deep learning is leading to new areas in biomedical applications. Wang and colleagues summarize in this Review the recent development and future directions of deep neural networks for superior image quality in the tomographic imaging field.

In ultrasound (US) imaging, various adaptive beamforming methods have been proposed to improve the resolution and contrast-to-noise ratio of the delay and sum (DAS) beamformers. Unfortunately, they often require computationally expensive calculations and their performance degrades when the underlying model is not sufficiently accurate. Moreover, ultrasound images usually require various type of post filtration such as deblurring and despeckling, etc., which further increase the complexity of the system. Deep learning-based solutions provides a quick remedy to these issue; however, in the current technology, a separate beamformer should be trained and stored for each application, demanding significant scanner resources. To address this problem, here we propose a switchable deep beamformer that can produce various types of output such as DAS, speckle removal, deconvolution, etc., using a single network with a simple switch. In particular, the switch is implemented through Adaptive Instance Normalization (AdaIN) layers, so that distinct outputs can be generated by merely changing the AdaIN code. Experimental results using B-mode focused ultrasound confirm the efficacy of the proposed methods.

Echocardiography is a pivotal imaging tool for emergency medicine. Unfortunately, it suffers from poor image quality due to the intrinsic limitations of sonography systems. Towards this end, a better quality can be achieved at the cost of reduced frame rate by increasing the number of transmit/receive events and utilizing computationally expensive noise suppression algorithms. However, this visual quality and temporal resolution trade-off is a bottleneck for many echocardiography applications. Conventional acceleration methods, such as multi-line acquisition (MLA), work only for limited acceleration factors and produce blocking artifacts at a high frame rate. Accordingly, various machine learning algorithms have been designed to reduce blocking artifacts in MLA. These algorithms require access to either high-quality raw RF data or time-delayed baseband IQ data. Unfortunately, in many lower-end commercial systems, such data are not accessible. On the other hand, ultrasound images are badly affected by speckle noises which significantly reduces the image quality. We propose an image domain unsupervised deep learning framework using cycleGAN architecture for high quality accelerated echocardiography that simultaneously reduces the blocking artifacts and the speckle noise. The method is evaluated on real in-vivo and phantom data and achieves notable performance gain.

In portable, three dimensional, and ultra-fast ultrasound imaging systems, there is an increasing demand for the reconstruction of high quality images from a limited number of radio-frequency (RF) measurements due to receiver (Rx) or transmit (Xmit) event sub-sampling. However, due to the presence of side lobe artifacts from RF sub-sampling, the standard beamformer often produces blurry images with less contrast, which are unsuitable for diagnostic purposes. Existing compressed sensing approaches often require either hardware changes or computationally expensive algorithms, but their quality improvements are limited. To address this problem, here we propose a novel deep learning approach that directly interpolates the missing RF data by utilizing redundancy in the Rx-Xmit plane. Our extensive experimental results using sub-sampled RF data from a multi-line acquisition B-mode system confirm that the proposed method can effectively reduce the data rate without sacrificing image quality., The title has been changed. This version will appear in IEEE Trans. on Medical Imaging

Image registration is a fundamental task in medical image analysis. Recently, many deep learning based image registration methods have been extensively investigated due to their comparable performance with the state-of-the-art classical approaches despite the ultra-fast computational time. However, the existing deep learning methods still have limitations in the preservation of original topology during the deformation with registration vector fields. To address this issues, here we present a cycle-consistent deformable image registration, dubbed CycleMorph. The cycle consistency enhances image registration performance by providing an implicit regularization to preserve topology during the deformation. The proposed method is so flexible that it can be applied for both 2D and 3D registration problems for various applications, and can be easily extended to multi-scale implementation to deal with the memory issues in large volume registration. Experimental results on various datasets from medical and non-medical applications demonstrate that the proposed method provides effective and accurate registration on diverse image pairs within a few seconds. Qualitative and quantitative evaluations on deformation fields also verify the effectiveness of the cycle consistency of the proposed method.

Cone-beam computed tomography (CBCT) used in radiotherapy (RT) has the advantage of being taken daily, but is difficult to use for purposes other than patient setup because of the poor image quality compared to fan-beam computed tomography (CT). Even though several methods have been proposed including the deformable image registration method to improve the quality of CBCT, the outcomes have not yet been satisfactory. Recently, deep learning has shown to produce high-quality results for various image-to-image translation tasks, suggesting the possibility of being an effective tool for converting CBCT into CT. In the field of RT, however, it may not always be possible to obtain paired datasets which consist of exactly matching CBCT and CT images. This study aimed to develop a novel, unsupervised deep-learning algorithm, which requires only unpaired CBCT and fan-beam CT images to remove the cone-beam artifact and thereby improve the quality of CBCT. Specifically, two cycle consistency generative adversarial networks (CycleGAN) were trained in the sagittal and coronal directions, and the generated results along those directions were then combined using spectral blending technique. To evaluate our methods, we applied it to American Association of Physicists in Medicine dataset. The experimental results show that our method outperforms the existing CycleGAN-based method both qualitatively and quantitatively.

Magnetic resonance imaging (MRI) acquisition is an inherently slow process whose acceleration has been the subject of much investigation. In recent years, the explosive advance of deep learning techniques for computer vision and image reconstruction has led to the investigation of deep neural networks for the reconstruction of MRI with under-sampled k-space. In this work, we propose a new image domain architecture that directly produces a sum-of-squares image from under-sampled multi-coil MRI acquisition. This model, called BarbellNet, is a fully convolutional neural network architecture that utilizes the channel attention mechanism using the residual channel attention block (RCAB). Through extensive experiments with the fastMRI data set, we confirm the efficacy of BarbellNet.

This paper reviews the NTIRE 2020 challenge on perceptual extreme super-resolution with focus on proposed solutions and results. The challenge task was to super-resolve an input image with a magnification factor 16 based on a set of prior examples of low and corresponding high resolution images. The goal is to obtain a network design capable to produce high resolution results with the best perceptual quality and similar to the ground truth. The track had 280 registered participants, and 19 teams submitted the final results. They gauge the state-of-the-art in single image super-resolution., Comment: CVPRW 2020

This paper reviews the second AIM learned ISP challenge and provides the description of the proposed solutions and results. The participating teams were solving a real-world RAW-to-RGB mapping problem, where to goal was to map the original low-quality RAW images captured by the Huawei P20 device to the same photos obtained with the Canon 5D DSLR camera. The considered task embraced a number of complex computer vision subtasks, such as image demosaicing, denoising, white balancing, color and contrast correction, demoireing, etc. The target metric used in this challenge combined fidelity scores (PSNR and SSIM) with solutions’ perceptual results measured in a user study. The proposed solutions significantly improved the baseline results, defining the state-of-the-art for practical image signal processing pipeline modeling.

BACKGROUND Cone-beam computed tomography (CBCT) is widely used in various medical imaging applications, including dental examinations. Dental CBCT images often suffer from motion artifacts caused by involuntary rigid motion of patients. However, earlier motion compensation studies are not applicable for dental CBCT systems using truncated detectors. OBJECTIVE This study proposes a novel motion correction algorithm that can be applied for truncated dental CBCT images. METHODS We propose a two-step method for motion correction. First, we estimate the relative displacement of each pair of opposite projections by finding the motion vector that maximizes the two-dimensional correlation coefficients of the opposite projections. Second, we convert the relative displacement into the absolute coordinate motion that yields the highest image sharpness of the reconstruction image. Using the motion vectors in the absolute coordinate system, motion artifacts are then compensated by modifying the trajectory of the source and detector during the back-projection step of the image reconstruction process. RESULTS In simulation, the proposed method successfully estimated the true relative displacement. After converting to the absolute coordinate motions, the motion-compensated image was close to the ground-truth image and exhibited a lower mean-square-error than that of the uncompensated image. The results from the real data experiment also confirmed that the proposed method successfully compensated for the motion artifacts. CONCLUSIONS The experimental results confirmed that the proposed method was applicable to most dental CBCT systems using a truncated detector without any use of an additional motion tracking system nor prior knowledge.

Diffuse optical tomography (DOT) has been investigated as an alternative imaging modality for breast cancer detection thanks to its excellent contrast to hemoglobin oxidization level. However, due to the complicated non-linear photon scattering physics and ill-posedness, the conventional reconstruction algorithms are sensitive to imaging parameters such as boundary conditions. To address this, here we propose a novel deep learning approach that learns non-linear photon scattering physics and obtains an accurate three dimensional (3D) distribution of optical anomalies. In contrast to the traditional black-box deep learning approaches, our deep network is designed to invert the Lippman-Schwinger integral equation using the recent mathematical theory of deep convolutional framelets. As an example of clinical relevance, we applied the method to our prototype DOT system. We show that our deep neural network, trained with only simulation data, can accurately recover the location of anomalies within biomimetic phantoms and live animals without the use of an exogenous contrast agent., Accepted for IEEE Trans. on Medical Imaging

Magnetic resonance imaging (MRI) is an inherently slow imaging modality, since it acquires multi-dimensional k-space data through 1-D free induction decay or echo signals. This often limits the use of MRI, especially for high resolution or dynamic imaging. Accordingly, many investigators has developed various acceleration techniques to allow fast MR imaging. For the last two decades, one of the most important breakthroughs in this direction is the introduction of compressed sensing (CS) that allows accurate reconstruction from sparsely sampled k-space data. The recent FDA approval of compressed sensing products for clinical scans clearly reflect the maturity of this technology. Therefore, this paper reviews the basic idea of CS and how this technology have been evolved for various MR imaging problems.

Medical image registration is one of the key processing steps for biomedical image analysis such as cancer diagnosis. Recently, deep learning based supervised and unsupervised image registration methods have been extensively studied due to its excellent performance in spite of ultra-fast computational time compared to the classical approaches. In this paper, we present a novel unsupervised medical image registration method that trains deep neural network for deformable registration of 3D volumes using a cycle-consistency. Thanks to the cycle consistency, the proposed deep neural networks can take diverse pair of image data with severe deformation for accurate registration. Experimental results using multiphase liver CT images demonstrate that our method provides very precise 3D image registration within a few seconds, resulting in more accurate cancer size estimation.

In ultrasound (US) imaging, individual channel RF measurements are back-propagated and accumulated to form an image after applying specific delays. While this time reversal is usually implemented using a hardware- or software-based delay-and-sum (DAS) beamformer, the performance of DAS decreases rapidly in situations where data acquisition is not ideal. Herein, for the first time, we demonstrate that a single data-driven adaptive beamformer designed as a deep neural network can generate high quality images robustly for various detector channel configurations and subsampling rates. The proposed deep beamformer is evaluated for two distinct acquisition schemes: focused ultrasound imaging and planewave imaging. Experimental results showed that the proposed deep beamformer exhibit significant performance gain for both focused and planar imaging schemes, in terms of contrast-to-noise ratio and structural similarity.

Purpose Magnetic resonance imaging (MRI) artifacts are originated from various sources including instability of an magnetic resonance (MR) system, patient motion, inhomogeneities of gradient fields, and so on. Such MRI artifacts are usually considered as irreversible, so additional artifact-free scan or navigator scan is necessary. To overcome these limitations, this article proposes a novel compressed sensing-based approach for removal of various MRI artifacts. Theory Recently, the annihilating filter based low-rank Hankel matrix approach was proposed. The annihilating filter based low-rank Hankel matrix exploits the duality between the low-rankness of weighted Hankel structured matrix and the sparsity of signal in a transform domain. Because MR artifacts usually appeared as sparse k-space components, the low-rank Hankel matrix from underlying artifact-free k-space data can be exploited to decompose the sparse outliers. Methods The sparse + low-rank decomposition framework using Hankel matrix was proposed for removal of MRI artifacts. Alternating direction method of multipliers algorithm was employed for the minimization of associated cost function with the initialized matrices from a factorization-based matrix completion. Results Experimental results demonstrated that the proposed algorithm can correct MR artifacts including herringbone (crisscross), motion, and zipper artifacts without image distortion. Conclusion The proposed method may be a robust correction solution for various MRI artifacts that can be represented as sparse outliers. Magn Reson Med 78:327-340, 2017. © 2016 International Society for Magnetic Resonance in Medicine.

Diffuse optical tomography (DOT) is a non-invasive imaging technique to reconstruct optical properties of biological tissues using near-infrared light, and it has been successfully used to measure functional brain activities via changes in cerebral blood volume and cerebral blood oxygenation. However, DOT presents a severely ill-posed inverse problem, so various types of regularization should be incorporated to overcome low spatial resolution and lack of depth sensitivity. Another limitation of the conventional DOT reconstruction methods is that an inference step is separately performed after the reconstruction, so complicated interaction between reconstruction and regularization is difficult to analyze. To overcome these technical difficulties, we propose a unified sparse recovery framework using a random effect model whose termination criterion is determined by the statistical inference. Both numerical and experimental results confirm that the proposed method outperforms the conventional approaches.

Dynamic positron emission tomography (PET) is widely used to measure changes in the bio-distribution of radiopharmaceuticals within particular organs of interest over time. However, to retain sufficient temporal resolution, the number of photon counts in each time frame must be limited. Therefore, conventional reconstruction algorithms such as the ordered subset expectation maximization (OSEM) produce noisy reconstruction images, thus degrading the quality of the extracted time activity curves (TACs). To address this issue, many advanced reconstruction algorithms have been developed using various spatio-temporal regularizations. In this paper, we extend earlier results and develop a novel temporal regularization, which exploits the self-similarity of patches that are collected in dynamic images. The main contribution of this paper is to demonstrate that the correlation of patches can be exploited using a low-rank constraint that is insensitive to global intensity variations. The resulting optimization framework is, however, non-Lipschitz and non-convex due to the Poisson log-likelihood and low-rank penalty terms. Direct application of the conventional Poisson image deconvolution by an augmented Lagrangian (PIDAL) algorithm is, however, problematic due to its large memory requirements, which prevents its parallelization. Thus, we propose a novel optimization framework using the concave-convex procedure (CCCP) by exploiting the Legendre?Fenchel transform, which is computationally efficient and parallelizable. In computer simulation and a real in vivo experiment using a high-resolution research tomograph (HRRT) scanner, we confirm that the proposed algorithm can improve image quality while also extracting more accurate region of interests (ROI) based kinetic parameters. Furthermore, we show that the total reconstruction time for HRRT PET is significantly accelerated using our GPU implementation, which makes the algorithm very practical in clinical environments.

This paper reviews the first challenge on single image super-resolution (restoration of rich details in an low resolution image) with focus on proposed solutions and results. A new DIVerse 2K resolution image dataset (DIV2K) was employed. The challenge had 6 competitions divided into 2 tracks with 3 magnification factors each. Track 1 employed the standard bicubic downscaling setup, while Track 2 had unknown downscaling operators (blur kernel and decimation) but learnable through low and high res train images. Each competition had ∽100 registered participants and 20 teams competed in the final testing phase. They gauge the state-of-the-art in single image super-resolution.

In portable, 3-D, or ultra-fast ultrasound (US) imaging systems, there is an increasing demand to reconstruct high quality images from limited number of data. However, the existing solutions require either hardware changes or computationally expansive algorithms. To overcome these limitations, here we propose a novel deep learning approach that interpolates the missing RF data by utilizing the sparsity of the RF data in the Fourier domain. Extensive experimental results from sub-sampled RF data from a real US system confirmed that the proposed method can effectively reduce the data rate without sacrificing the image quality., Comment: Invited paper for ICASSP 2018 Special Session for "Machine Learning in Medical Imaging: from Measurement to Diagnosis"

Recently, an annihilating filter based low-rank Hankel matrix approach (ALOHA) was proposed as a general framework for sparsity-driven k-space interpolation method for compressed sensing MRI (CS-MRI). The principle of ALOHA framework is based on the fundamental duality between the transform domain sparsity in the primary space and the low-rankness of weighted Hankel matrix in Fourier domain, which converts CS-MRI to a k-space interpolation problem using structured matrix completion. In this review, we explain the theory behind ALOHA. Experimental results with in vivo data for multi-coil dynamic imaging, parametric mapping as well as Nyquist ghost correction confirmed that the proposed method has potential to be a general solution of various MR imaging problems.

To reduce data rate for power-limited portable ultrasound imaging systems, various compressed sensing approaches have been investigated. However, most of the existing approaches require either hardware changes or computationally expensive forward modeling of wave propagation. To overcome these limitations, here we propose a novel low rank interpolation method that exploits the annihilation property of ultrasound measurements. Specifically, based on the observation that the pre-beamformed signals from a ultrasound probe have significant redundancy in spatial domain, only random subset of scan line measurements are obtained and the missing data are interpolated using a recently proposed annihilating filter-based low-rank Hankel matrix approach (ALOHA). Moreover, we exploit temporal correlation between the measurements from consecutive frames by augmenting Hankel matrix side by side for a low rank matrix completion. Reconstruction results confirmed that the proposed method can effectively reduce the data rate for ultrasound acquisition without sacrificing the image quality.

Dynamic contrast enhanced (DCE) MRI is one of the important tools for diagnosis. In particular, time-resolved angiography with interleaved stochastic trajectories (TWIST) has been widely used thanks to the improved temporal and spatial resolution. However, the temporal resolution of TWIST is not a true one, because the periphery of k-space data from several time frames should be shared to produce one image. To address this issue, here we proposed a high performance low-rank k-space interpolation method that significantly improves the temporal resolution without sacrificing the image quality. The main idea of the proposed method is originated from the state-of-art MR reconstruction methods using annihilating filter-based low rank Hankel matrix approach (ALOHA). Experimental results using in-vivo data showed significantly improved temporal resolution compared with the standard reconstruction method.

In this paper, I will present an annihilating filter-based low-rank Hankel matrix approach for biomedical imaging and image processing that converts a compressive image reconstruction problem as a low-rank interpolation problems. The theoretical performance guarantee as well as various applications in imaging and image processing will be demonstrated.

Diffuse optical tomography (DOT) is a sensitive and relatively low cost imaging modality that reconstructs optical properties of a highly scattering medium. However, due to the diffusive nature of light propagation, the problem is severely ill-conditioned and highly nonlinear. Even though nonlinear iterative methods have been commonly used, they are computationally expensive especially for three dimensional imaging geometry. Recently, compressed sensing theory has provided a systematic understanding of high resolution reconstruction of sparse objects in many imaging problems; hence, the goal of this paper is to extend the theory to the diffuse optical tomography problem. The main contributions of this paper are to formulate the imaging problem as a joint sparse recovery problem in a compressive sensing framework and to propose a novel noniterative and exact inversion algorithm that achieves the l(0) optimality as the rank of measurement increases to the unknown sparsity level. The algorithm is based on the recently discovered generalized MUSIC criterion, which exploits the advantages of both compressive sensing and array signal processing. A theoretical criterion for optimizing the imaging geometry is provided, and simulation results confirm that the new algorithm outperforms the existing algorithms and reliably reconstructs the optical inhomogeneities when we assume that the optical background is known to a reasonable accuracy.

A compressed sensing dynamic MR technique called k-t FOCUSS (k-t FOCal Underdetermined System Solver) has been recently proposed. It outperforms the conventional k-t BLAST/SENSE (Broad-use Linear Acquisition Speed-up Technique/SENSitivity Encoding) technique by exploiting the sparsity of x-f signals. This paper applies this idea to radial trajectories for high-resolution cardiac cine imaging. Radial trajectories are more suitable for high-resolution dynamic MRI than Cartesian trajectories since there is smaller tradeoff between spatial resolution and number of views if streaking artifacts due to limited views can be resolved. As shown for Cartesian trajectories, k-t FOCUSS algorithm efficiently removes artifacts while preserving high temporal resolution. k-t FOCUSS algorithm applied to radial trajectories is expected to enhance dynamic MRI quality. Rather than using an explicit gridding method, which transforms radial k-space sampling data to Cartesian grid prior to applying k-t FOCUSS algorithms, we use implicit gridding during FOCUSS iterations to prevent k-space sampling errors from being propagated. In addition, motion estimation and motion compensation after the first FOCUSS iteration were used to further sparsify the residual image. By applying an additional k-t FOCUSS step to the residual image, improved resolution was achieved. In vivo experimental results show that this new method can provide high spatiotemporal resolution even from a very limited radial data set.

A model-based dynamic MRI called k-t BLAST/SENSE has drawn significant attention from the MR imaging community because of its improved spatio-temporal resolution. Recently, we showed that the k-t BLAST/SENSE corresponds to the special case of a new dynamic MRI algorithm called k-t FOCUSS that is optimal from a compressed sensing perspective. The main contribution of this article is an extension of k-t FOCUSS to a more general framework with prediction and residual encoding, where the prediction provides an initial estimate and the residual encoding takes care of the remaining residual signals. Two prediction methods, RIGR and motion estimation/compensation scheme, are proposed, which significantly sparsify the residual signals. Then, using a more sophisticated random sampling pattern and optimized temporal transform, the residual signal can be effectively estimated from a very small number of k-t samples. Experimental results show that excellent reconstruction can be achieved even from severely limited k-t samples without aliasing artifacts. Magn Reson Med 61:103–116, 2009.

The combination of the PI with CS reconstruction has been an important research topic in MR community due to its potential for further acceleration in MR acquisition. In particular, the application to the 3D volumetric imaging has great potential due to the introduction of additional dimension along slice direction. In this work, we propose a combined approach of dynamic compressed sensing (CS) and parallel imaging (PI) for dynamic simultaneous multislice (SMS) imaging using multi-band RF pulses. The proposed algorithm is tested to dynamic contrast-enhanced angiography to observe brain tissue dynamics. SMS image of slice-direction acceleration factor of 2,4 and 8 with in-plane acceleration of 2 are reconstructed by the proposed multi-band dynamic CS algorithm. The reconstructed results show that a proper combination of CS and parallel imaging can effectively suppress noise propagation and provide improved spatio-temporal resolution under high acceleration factor.

Spectral computed tomography (CT) is a promising technique with the potential for improving lesion detection, tissue characterization, and material decomposition. In this paper, we are interested in kVp switching-based spectral CT that alternates distinct kVp X-ray transmissions during gantry rotation. This system can acquire multiple X-ray energy transmissions without additional radiation dose. However, only sparse views are generated for each spectral measurement; and the spectra themselves are limited in number. To address these limitations, we propose a penalized maximum likelihood method using spectral patch-based low-rank penalty, which exploits the self-similarity of patches that are collected at the same position in spectral images. The main advantage is that the relatively small number of materials within each patch allows us to employ the low-rank penalty that is less sensitive to intensity changes while preserving edge directions. In our optimization formulation, the cost function consists of the Poisson log-likelihood for X-ray transmission and the nonconvex patch-based low-rank penalty. Since the original cost function is difficult to minimize directly, we propose an optimization method using separable quadratic surrogate and concave convex procedure algorithms for the log-likelihood and penalty terms, which results in an alternating minimization that provides a computational advantage because each subproblem can be solved independently. We performed computer simulations and a real experiment using a kVp switching-based spectral CT with sparse-view measurements, and compared the proposed method with conventional algorithms. We confirmed that the proposed method improves spectral images both qualitatively and quantitatively. Furthermore, our GPU implementation significantly reduces the computational cost.

One of important applications of dynamic contrast-enhanced (DCE) MRI is to observe brain tissue dynamics. In this work, we applied a combined approach of dynamic compressed sensing (CS) and parallel imaging for volumetric DCE-MRI. The combination of the parallel imaging with CS reconstruction has been a topic of significant interest in MR community due to its potential for further acceleration in MR acquisition. In particular, the application to the 3D volumetric imaging has great potential due to the introduction of additional dimension along slice direction. Here, we proposed two different combinations depending on their encoding schemes: i.e. 3D encoding and simultaneous multislice imaging (SMS) using multiband RF pulses. Experiments show that a proper combination of CS and parallel imaging can effectively suppress noise propagation and provide improved spatio-temporal resolution.

One of the technical challenges in cine magnetic resonance imaging (MRI) is to reduce the acquisition time to enable the high spatio-temporal resolution imaging of a cardiac volume within a short scan time. Recently, compressed sensing approaches have been investigated extensively for highly accelerated cine MRI by exploiting transform domain sparsity using linear transforms such as wavelets, and Fourier. However, in cardiac cine imaging, the cardiac volume changes significantly between frames, and there often exist abrupt pixel value changes along time. In order to effectively sparsify such temporal variations, it is necessary to exploit temporal redundancy along motion trajectories. This paper introduces a novel patch-based reconstruction method to exploit geometric similarities in the spatio-temporal domain. In particular, we use a low rank constraint for similar patches along motion, based on the observation that rank structures are relatively less sensitive to global intensity changes, but make it easier to capture moving edges. A Nash equilibrium formulation with relaxation is employed to guarantee convergence. Experimental results show that the proposed algorithm clearly reconstructs important anatomical structures in cardiac cine image and provides improved image quality compared to existing state-of-the-art methods such as k-t FOCUSS, k-t SLR, and MASTeR.

Localization microscopy achieves nanoscale spatial resolution by iterative localization of sparsely activated molecules, which generally leads to a long acquisition time. By implementing advanced algorithms to treat overlapping point spread functions (PSFs), imaging of densely activated molecules can improve the limited temporal resolution, as has been well demonstrated in two-dimensional imaging. However, three-dimensional (3D) localization of high-density data remains challenging since PSFs are far more similar along the axial dimension than the lateral dimensions. Here, we present a new, high-density 3D imaging system and algorithm. The hybrid system is implemented by combining astigmatic and biplane imaging. The proposed 3D reconstruction algorithm is extended from our state-of-the art 2D high-density localization algorithm. Using mutual coherence analysis of model PSFs, we validated that the hybrid system is more suitable than astigmatic or biplane imaging alone for 3D localization of high-density data. The efficacy of the proposed method was confirmed via simulation and real data of microtubules. Furthermore, we also successfully demonstrated fluorescent-protein-based live cell 3D localization microscopy with a temporal resolution of just 3 seconds, capturing fast dynamics of the endoplasmic recticulum. (C) 2014 Optical Society of America

Dynamic 4D x-ray computed tomography (CT) is often used for several applications such as respiratory and cardiac imaging. To acquire enough data for several phases, repeated x-ray CT scans should be conducted. Thus, the dose reduction is one of important issues for dynamic CT scans. Another issue is that insufficient angular sampling can be occurred by fast motions such as a cardiac CT imaging. In this paper, the main goal is to develop a patch based low rank regularization approach that exploits inherent similarities between dynamic images for reducing both noise and motion artifacts incurred by low dose and insufficient angular sampling, respectively. In particular, our optimization framework is based on the Poisson log-likelihood and the non-convex low rank regularization. To address the non-separable and non-convex optimization problem, we used the separable paraboloidal surrogates (SPS) for Poisson log-likelihood, and the concave-convex procedure (CCCP) for non-convex low rank regularization, which guarantees the fast convergence.We confirm that the proposed algorithm can provide significantly improved image quality.

In this paper, a novel patch-based signal processing algorithm for motion estimated/compensated compressed sensing dynamic MR imaging is proposed. More specifically, we impose a non-convex patch-based low-rank penalty that exploits self-similarities within the images. This penalty is shown to favor capturing geometric features such as edges rather than reconstructing the background noises. To solve the resulting non-convex optimization problem, we propose a globally convergent concave-convex procedure (CCCP) using convex conju- gate, which has closed form solution at each sub-iteration. Experimental results demonstrate that the proposed algorithm outperforms the existing ones.

There is increasing demand in the field of dental and medical radiography for effective metal artifact reduction (MAR) in computed tomography (CT) because artifact caused by metallic objects causes serious image degradation that ob- scures information regarding the teeth and/or other biological structures. This paper presents a new MAR method that uses the Laplacian operator to reveal background projection data hidden in regions containing data from metal. In the proposed method, we attempted to decompose the projection data into two parts: data from metal only (metal data), and background data in the absence of metal. Removing metal data from the projections enables us to perform sparsity-driven reconstruction of the metal component and subsequent removal of the metal artifact. The results of clinical experiments demonstrated that the proposed MAR algorithm improves image quality and increases the standard of 3D reconstruction images of the teeth and mandible.

Recently, it has been shown that interior tomography problems in x-ray CT can be uniquely determined if tiny subregions inside of the region of interest are known. The solution can be obtained by the projection onto convex sets (POCS) combined with the backprojection filtration algorithm. However, it is well-known that the convergence speed of POCS is slow; hence, to overcome the limitation, this paper employs a parallel proximal algorithm (PPXA) to simultaneously consider multiple convex constraints rather than projecting on each of them sequentially as in POCS. Our simulation results show that the solution for the interior tomography problem can be accurately obtained using PPXA with a much smaller number of iterations than POCS.

Super-resolution localization microscopy relies on sparse activation of photo-switchable probes. Such activation, however, introduces limited temporal resolution. High-density imaging overcomes this limitation by allowing several neighboring probes to be activated simultaneously. In this work, we propose an algorithm that incorporates a continuous-domain sparsity prior into the high-density localization problem. We use a Taylor approximation of the PSF, and rely on a fast proximal gradient optimization procedure. Unlike currently available methods that use discrete-domain sparsity priors, our approach does not restrict the estimated locations to a pre-defined sampling grid. Experimental results of simulated and real data demonstrate significant improvement over these methods in terms of accuracy, molecular identification and computational complexity.

Localization of brain signal sources from EEG/MEG has been an active area of research [1]. Currently, there exists a variety of approaches such as MUSIC [2], M-SBL [3], and etc. These algorithms have been applied for various clinical examples and demonstrated excellent performances. However, when the unknown sources are highly correlated, the conventional algorithms often exhibit spurious reconstructions. To address the problem, this paper proposes a new algorithm that generalizes M-SBL by exploiting the fundamental subspace geometry in the multiple measurement problem (MMV). Experimental results using simulation and real phantom data show that the proposed algorithm outperforms the existing methods even under a highly correlated source condition.

A variety of far-field super-resolution microscopy techniques have been proposed recently by exploiting nonlinear phenomenon in optics or specialized photo-switchable probes. Here, we present a linear optics based super-resolution microscopy for standard probes and experimental protocols. Rather than exploiting the nonlinearity in optics or photo-switchable probes, we use a nonlinear reconstruction algorithm that exploits joint sparsity of fluorescent emission from multiple random illumination. To maximize the resolution improvement owing to the joint sparsity, spatially incoherent speckle illumination is used for conventional epi-fluorescence microscopy setup. Experimental results obtained from a nano-pattern and bio-samples stained with standard fluorescent probes along with the proposed method demonstrate that a resolution of up to 37nm can be achieved.

Functional near infrared spectroscopy (fNIRS) is a non-invasive imaging modality to measure functional brain activities. Many researches have investigated diffuse optical tomography (DOT) to overcome the limitation of lack of depth information in fNIRS topographic approach. In this paper, we proposes a novel compressed sensing approach, especially using a 2-thresholding algorithm, that directly reconstructs statistical parameter maps by exploiting spatio-temporal sparsity of neural activation. The final reconstruction algorithm has very intuitive form which is similar to conventional fDOT with SPM analysis. However, the main advantage of the new algorithm is that the unknown weighting components in inversion kernel are iteratively updated for more accurate reconstruction which significantly improves the reconstruction performance. Experimental results demonstrated that the localization error using the proposed method is competitive with that of fMRI.

A non-local motion compensation algorithm combined with k-t FOCUSS for high resolution compressed sensing cardiac cine imaging is proposed in this paper. While conventional non-local means algorithms use self-similarity, our method collects similar blocks from another reference frame that is obtained during diastole phase. We show that this non-local motion compensation is optimal in MMSE sense, if the probability for each candidate patch is uniform. Experimental results show that the proposed algorithm clearly reconstructs important edge boundaries in cardiac cine imaging and provides significant performance improvement compared to the existing motion compensated k-t FOCUSS.

In digital tomosynthesis (DTS), accurate scatter correction is often necessary for quantitative analysis. This is especially important because low energy x-ray of 10–40 keV, which is widely used for the breast imaging to enhance the contrast between adipose and glandular, results in high scatter fraction. In this paper, we propose an iterative scatter correction for digital tomopsynthesis using composition ratio update and GPU based Monte Carlo simulation (MCS). One of the technical difficulty in scatter estimation using MCS for tomosynthesis is that accurate segmentation of 3D volume is very difficult due to the low resolution of the reconstruction object. Thus, an intermediate surrogate object is introduced to represent composition ratio between adipose and glandular. We show that the composition ratio can be calculated using average attenuation coefficients. Another technical challenge is extremely high computational cost of MCS. We overcome this using GPU based ultra-fast MCS. Our results demonstrate that our iterative scatter correction using composition ratio update is indeed effective in improving the quality of the reconstruction object in a reasonable time frame.

Accurate scatter correction is especially important for high-resolution 3D positron emission tomographies (PETs) such as high-resolution research tomograph (HRRT) due to large scatter fraction in the data. To address this problem, a fully 3D iterative scatter-corrected ordered subset expectation maximization (OSEM) in which a 3D single scatter simulation (SSS) is alternatively performed with a 3D OSEM reconstruction was recently proposed. However, due to the computational complexity of both SSS and OSEM algorithms for a high-resolution 3D PET, it has not been widely used in practice. The main objective of this paper is, therefore, to accelerate the fully 3D iterative scatter-corrected OSEM using a graphics processing unit (GPU) and verify its performance for an HRRT. We show that to exploit the massive thread structures of the GPU, several algorithmic modifications are necessary. For SSS implementation, a sinogram-driven approach is found to be more appropriate compared to a detector-driven approach, as fast linear interpolation can be performed in the sinogram domain through the use of texture memory. Furthermore, a pixel-driven backprojector and a ray-driven projector can be significantly accelerated by assigning threads to voxels and sinograms, respectively. Using Nvidia's GPU and compute unified device architecture (CUDA), the execution time of a SSS is less than 6 s, a single iteration of OSEM with 16 subsets takes 16 s, and a single iteration of the fully 3D scatter-corrected OSEM composed of a SSS and six iterations of OSEM takes under 105 s for the HRRT geometry, which corresponds to acceleration factors of 125× and 141× for OSEM and SSS, respectively. The fully 3D iterative scatter-corrected OSEM algorithm is validated in simulations using Geant4 application for tomographic emission and in actual experiments using an HRRT.

Diffuse optical tomography (DOT) is an emerging imaging modality to reconstruct the optical parameters of a highly scattering medium. However, ill-posedness and nonlinearity of light scattering make the DOT problem very difficult. We showed that DOT problem can be formulated as a multiple measurement vector (MMV) problem, and compressive sensing approach like S-OMP can be used. However, for a limited number of illumination patterns, the conventional compressed sensing (CS) approach for DOT was not satisfactory. The main objective of this paper is to propose a new non-iterative and exact reconstruction algorithm for the diffuse optical tomography problem that outperforms the conventional compressive sensing approach, thanks to a recently invented generalized MUSIC algorithm. Simulation results confirm that the new algorithm outperforms the previous algorithms and reliably reconstructs optical inhomogeneities.