Transformers have made remarkable progress towards modeling long-range dependencies within the medical image analysis domain. However, current transformer-based models suffer from several disadvantages: (1) existing methods fail to capture important features of images due naive tokenization scheme; (2) information loss because they only consider single-scale feature representations; and (3) segmentation label maps generated by are not accurate enough without considering rich semantic...

Limited-angle X-ray computed tomography (CT) reconstruction is a typical ill-posed problem. To recover satisfied reconstructed images with limited-angle CT projections, prior information usually introduced into image reconstruction, such as the piece-wise constant, nonlocal similarity, and so on. further improve quality for dictionary learning (DL) gradient ℓ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sub> -norm are combined model, it...

Spectral computed tomography (CT) reconstructs multienergy images from data in different energy bins. However, these reconstructed can be contaminated by noise due to the limited numbers of photons corresponding In this paper, we propose a spectral CT reconstruction method aided self-similarity image-spectral tensors, which utilizes selfsimilarity patches both spatial and domains. Patches with similar structures identified joint searching strategy form basic tensor unit, utilized improve...

With the development of low-dose computed tomography (CT), incomplete data reconstruction has been widely concerned. The total variation (TV) minimization algorithm can accurately reconstruct images from sparse or noisy data.However, traditional TV ignores direction structures in images, leading to loss edge information and block artifacts when object is not piecewise constant. Since anisotropic facilitate preserving detail we aim improve terms accuracy via this approach.In paper, propose an...

The adoption of large language models (LLMs) in healthcare has garnered significant research interest, yet their performance remains limited due to a lack domain‐specific knowledge, medical reasoning skills, and unimodal nature, which restricts them text‐only inputs. To address these limitations, we propose MultiMedRes, multimodal collaborative framework that simulates human physicians’ communication by incorporating learner agent proactively acquire information from expert models....

Background: Orthogonal translation computed laminography (OTCL) has great potential for tiny fault detection in laminated structure thin-plate parts. It offers a larger magnification ratio but generates limited projection data, which would result aliasing artifacts the reconstructed image. Objective: One way to minimize these is use prior information, such as piecewise constant property and image information. This work was inspired by adaptive-weighted high order total variation (awHOTV)...

Abstract The photon-counting detector based spectral computed tomography (CT) is promising for lesion detection, tissue characterization, and material decomposition. However, the lower signal-to-noise ratio within multi-energy projection dataset can result in poorly reconstructed image quality. Recently, as prior information, a high-quality mean was introduced into constrained compressed sensing (PICCS) framework to suppress noise, leading PICCS (SPICCS). In original SPICCS model, gradient L...

The spectral computed tomography (CT) has huge advantages by providing accurate material information. Unfortunately, due to the instability or overdetermination of decomposition model, accuracy can be compromised in practice. Very recently, dictionary learning based image-domain (DLIMD) obtain high for decompositions from reconstructed CT images. This method explore correlation components some extent training a unified all In addition, prior as penalty is applied on independently, and many...

Spectral-computed tomography (CT) has been demonstrating its great advantages in lesion detection, tissue characterization, and material decomposition. However, the quality of images is often significantly corrupted with various noises, which brings a challenge for applications. Because channel-wise from different energy interval share similar structure physical message, spatial sparsity, global correlation across spectrum (GCS), nonlocal self-similarity (NSS) as three important...

Spectral computed tomography (CT) reconstructs the same scanned object from projections of multiple narrow energy windows, and it can be used for material identification decomposition. However, multi-energy projection dataset has a lower signal-noise-ratio (SNR), resulting in poor reconstructed image quality. To address this thorny problem, we develop spectral CT reconstruction method, namely spatial-spectral cube matching frame (SSCMF). This method is inspired by following three facts: i)...

Spectral computed tomography (CT) reconstructs material-dependent attenuation images from the projections of multiple narrow energy windows, which is meaningful for material identification and decomposition. Unfortunately, multi-energy projection datasets usually have lower signal-noise ratios (SNR). Very recently, a spatial-spectral cube matching frame (SSCMF) was proposed to explore non-local similarities spectral CT. This method constructs group by clustering up series cubes. The small...

Micro computed tomography (µCT) allows the noninvasive visualization and 3D reconstruction of internal structures objects with high resolution. However, current commercial µCT system relatively rotates source-detector or to collect projections, referred as RCT in this paper, has difficulties imaging large resolutions because fabrication large-area, inexpensive flat-panel detectors remains a challenge. In we proposed source translation based CT (STCT) for resolution get rid limitation...

Recent studies on contrastive learning have achieved remarkable performance solely by leveraging few labels in the context of medical image segmentation. Existing methods mainly focus instance discrimination and invariant mapping. However, they face three common pitfalls: (1) tailness: data usually follows an implicit long-tail class distribution. Blindly all pixels training hence can lead to imbalance issues, cause deteriorated performance; (2) consistency: it remains unclear whether a...

The spectral computed tomography (CT) system based on a photon-counting detector (PCD) can quantitatively analyze the material composition of inspected object by decomposition. Nonetheless, raw projection CT is frequently disturbed noise and artifacts, resulting in poor quality decomposition images. Recently, generalized dictionary learning image-domain (GDLIMD) to obtain high-quality DL has great advantages suppression while its protection fine structure edge information insufficient. To...

Micro-computed tomography (micro-CT) is an indispensable tool to provide attenuation-based, high-resolution 3D images in scientific research. However, its current available configuration limits the size of objects that can be imaged. Previously, we have proposed a multiple source translation computed (mSTCT) imaging geometry extend field-of-view micro-CT, and developed corresponding reconstruction algorithm called virtual projection-based filtered back-projection (V-FBP). V-FBP achieves...