Optical Coherence Tomography (OCT) is a non-invasive technique for obtaining detailed, cross-sectional images of coronary arteries. However, cost-effective OCT systems produce only low-resolution (LR) images. Unsuperv...
详细信息
ISBN:
(数字)9798350344813
ISBN:
(纸本)9798350344820
Optical Coherence Tomography (OCT) is a non-invasive technique for obtaining detailed, cross-sectional images of coronary arteries. However, cost-effective OCT systems produce only low-resolution (LR) images. Unsupervised OCT super-resolution (OCT-SR) presents a cost-effective solution, eliminating the need for high-resolution (HR) systems or co-registered LR-HR image pairs. Existing unsupervised OCT-SR methods formulate the SR task as an image-to-image translation problem, and use CycleGAN as their backbone. However, CycleGAN is known to lack translation identifiability that can result in incorrect SR results. Existing methods often empirically combat this issue by using multiple regularization terms to incorporate expert-annotated side information, resulting in complicated learning losses and extensive annotations. This work proposes a translation identifiability-guided framework based on recent advances in unsupervised domain translation. Employing a diversified distribution matching module, our approach guarantees OCT translation identifiability under reasonable conditions using a simple and succinct learning loss. Numerical results indicate that our framework matches or surpasses the state-of-the-art (SOTA) baseline's performance while requiring considerably fewer resources, e.g., annotations, computation time, and memory.
Photoacoustic imaging provides a high resolution imaging solution for biological tissues. In this paper, we consider recovering the z-averaged spatial distribution using a cylindrical array of sensors long in the z-di...
详细信息
Photoacoustic imaging provides a high resolution imaging solution for biological tissues. In this paper, we consider recovering the z-averaged spatial distribution using a cylindrical array of sensors long in the z-direction. This configuration can be applied for breast imaging. In previous frequency domain algorithms there is no way to determine the optimum, minimum number of frequency samples required. We propose a novel frequency domain algorithm which uses a minimum number of frequency samples for image reconstruction. Further, we propose a method to calculate this minimum number of frequency samples. Provided there is no aliasing our proposed method, using a reduced number of frequency samples, provides similar results to previous frequency reconstruction methods.
Technological aspects of the 3D integration of a multilayer combined mixed-signal and digital sensor-processor array chip is described. The 3D integration raises the question of signal routing, power distribution, and...
详细信息
Technological aspects of the 3D integration of a multilayer combined mixed-signal and digital sensor-processor array chip is described. The 3D integration raises the question of signal routing, power distribution, and heat dissipation, which aspects are considered systematically in the digital processor array layer as part of the multi layer structure. We have developed a linear programming based evaluation system to identify the proper architecture and its parameters.
Massive multiple-input-multiple-output (MIMO) is a core technology of current and future wireless networks. However, the very large dimension of a massive antenna array can lead to radical changes in the electromagnet...
详细信息
ISBN:
(数字)9798350344813
ISBN:
(纸本)9798350344820
Massive multiple-input-multiple-output (MIMO) is a core technology of current and future wireless networks. However, the very large dimension of a massive antenna array can lead to radical changes in the electromagnetic fields near the array, and the classical far-field channel model is no longer accurate. Instead, the channel should be modeled under the assumption of near-field spherical wavefronts. Furthermore, the very large dimension of the arrays can also result in high power consumption and hardware complexity. A practical solution for this problem is to use low-resolution analog-to-digital converters (ADCs). It is therefore of significance to study the near-field channel estimation problem for MIMO systems implemented with low-resolution ADCs. We propose an efficient on-grid polar-domain channel estimation method which relies on the polar-domain sparsity of the near-field channels. We first reformulate the sparse low-resolution near-field maximum-likelihood channel estimation problem by exploiting an approximation of the cu-mulative distribution function of a normal random variable as a logistic activation function. We then develop an on-grid polar-domain channel estimation method based on the gradient descent approach and the polar-domain sparsity of the near-field channel. Finally, we apply the deep unfolding technique to optimize the performance of the proposed method and illustrate its efficiency via several simulation studies.
The detection of central apneas using an unobtrusive pressure sensorarray installed in the beds of smart homes could allow comfortable diagnosis of sleep disturbances. To improve central apnea detection, two methods ...
详细信息
The detection of central apneas using an unobtrusive pressure sensorarray installed in the beds of smart homes could allow comfortable diagnosis of sleep disturbances. To improve central apnea detection, two methods of improving the results of apneas classified by a previously developed method are presented: moving average windowing and window elimination. The first improved classifier sensitivity, while the second improved the specificity with better duration estimation. However, it was slightly more likely to miss apnea segments altogether.
Sparse subspace clustering (SSC) using greedy- based neighbor selection, such as matching pursuit (MP) and orthogonal matching pursuit (OMP), has been known as a popular computationally-efficient alternative to the co...
详细信息
ISBN:
(数字)9781728119465
ISBN:
(纸本)9781728119472
Sparse subspace clustering (SSC) using greedy- based neighbor selection, such as matching pursuit (MP) and orthogonal matching pursuit (OMP), has been known as a popular computationally-efficient alternative to the conventional ℓ 1 -minimization based solutions. Under deterministic bounded noise corruption, in this paper we derive coherence-based sufficient conditions guaranteeing correct neighbor identification using MP/OMP. Our analyses exploit the maximum/minimum inner product between two noisy data points subject to a known upper bound on the noise level. The obtained sufficient condition clearly reveals the impact of noise on greedy-based neighbor recovery. Specifically, it asserts that, as long as noise is sufficiently small and the resultant perturbed residual vectors stay close to the desired subspace, both MP and OMP succeed in returning a correct neighbor subset. Extensive numerical experiments are used to corroborate our theoretical study. A striking finding is that, as long as the ground truth subspaces are well-separated from each other, MP-based iterations, while enjoying lower algorithmic complexity, yields smaller perturbed residuals, thereby better able to identify correct neighbors and, in turn, achieving higher global data clustering accuracy.
暂无评论