Search

The segmentation of the orbit in computed tomography (CT) images plays a crucial role in facilitating the quantitative analysis of orbital decompression surgery for patients with Thyroid-associated Ophthalmopathy (TAO). However, the task of orbit segmentation, particularly in postoperative images, remains challenging due to the significant shape variation and limited amount of labeled data. In this paper, we present a two-stage semi-supervised framework for the automatic segmentation of the orbit in both preoperative and postoperative images, which consists of a pseudo-label generation stage and a semi-supervised segmentation stage. A Paired Copy-Paste strategy is concurrently introduced to proficiently amalgamate features extracted from both preoperative and postoperative images, thereby augmenting the network discriminative capability in discerning changes within orbital boundaries. More specifically, we employ a random cropping technique to transfer regions from labeled preoperative images (foreground) onto unlabeled postoperative images (background), as well as unlabeled preoperative images (foreground) onto labeled postoperative images (background). It is imperative to acknowledge that every set of preoperative and postoperative images belongs to the identical patient. The semi-supervised segmentation network (stage 2) utilizes a combination of mixed supervisory signals from pseudo labels (stage 1) and ground truth to process the two mixed images. The training and testing of the proposed method have been conducted on the CT dataset obtained from the Eye Hospital of Wenzhou Medical University. The experimental results demonstrate that the proposed method achieves a mean Dice similarity coefficient (DSC) of 91.92% with only 5% labeled data, surpassing the performance of the current state-of-the-art method by 2.4%., Competing Interests: Declaration of competing interest The authors declare that they have no competing interests., (Copyright © 2024. Published by Elsevier Ltd.)

Feature matching of monocular laparoscopic videos is crucial for visualization enhancement in computer-assisted surgery, and the keys to conducting high-quality matches are accurate homography estimation, relative pose estimation, as well as sufficient matches and fast calculation. However, limited by various monocular laparoscopic imaging characteristics such as highlight noises, motion blur, texture interference and illumination variation, most exiting feature matching methods face the challenges of producing high-quality matches efficiently and sufficiently. To overcome these limitations, this paper presents a novel sequential coupling feature descriptor to extract and express multilevel feature maps efficiently, and a dual-correlate optimized coarse-fine strategy to establish dense matches in coarse level and adjust pixel-wise matches in fine level. Firstly, a novel sequential coupling swin transformer layer is designed in feature descriptor to learn and extract multilevel feature representations richly without increasing complexity. Then, a dual-correlate optimized coarse-fine strategy is proposed to match coarse feature sequences under low resolution, and the correlated fine feature sequences is optimized to refine pixel-wise matches based on coarse matching priors. Finally, the sequential coupling feature descriptor and dual-correlate optimization are merged into the Sequential Coupling Dual-Correlate Network (SeCo DC-Net) to produce high-quality matches. The evaluation is conducted on two public laparoscopic datasets: Scared and EndoSLAM, and the experimental results show the proposed network outperforms state-of-the-art methods in homography estimation, relative pose estimation, reprojection error, matching pairs number and inference runtime. The source code is publicly available at https://github.com/Iheckzza/FeatureMatching., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023. Published by Elsevier Ltd.)

Automatic vessel segmentation is a critical area of research in medical image analysis, as it can greatly assist doctors in accurately and efficiently diagnosing vascular diseases. However, accurately extracting the complete vessel structure from images remains a challenge due to issues such as uneven contrast and background noise. Existing methods primarily focus on segmenting individual pixels and often fail to consider vessel features and morphology. As a result, these methods often produce fragmented results and misidentify vessel-like background noise, leading to missing and outlier points in the overall segmentation. To address these issues, this paper proposes a novel approach called the progressive edge information aggregation network for vessel segmentation (PEA-Net). The proposed method consists of several key components. First, a dual-stream receptive field encoder (DRE) is introduced to preserve fine structural features and mitigate false positive predictions caused by background noise. This is achieved by combining vessel morphological features obtained from different receptive field sizes. Second, a progressive complementary fusion (PCF) module is designed to enhance fine vessel detection and improve connectivity. This module complements the decoding path by combining features from previous iterations and the DRE, incorporating nonsalient information. Additionally, segmentation-edge decoupling enhancement (SDE) modules are employed as decoders to integrate upsampling features with nonsalient information provided by the PCF. This integration enhances both edge and segmentation information. The features in the skip connection and decoding path are iteratively updated to progressively aggregate fine structure information, thereby optimizing segmentation results and reducing topological disconnections. Experimental results on multiple datasets demonstrate that the proposed PEA-Net model and strategy achieve optimal performance in both pixel-level and topology-level metrics., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023 Elsevier Ltd. All rights reserved.)

Background: Monocular depth estimation plays a fundamental role in clinical endoscopy surgery. However, the coherent illumination, smooth surfaces, and texture-less nature of endoscopy images present significant challenges to traditional depth estimation methods. Existing approaches struggle to accurately perceive depth in such settings., Method: To overcome these challenges, this paper proposes a novel multi-scale residual fusion method for estimating the depth of monocular endoscopy images. Specifically, we address the issue of coherent illumination by leveraging image frequency domain component space transformation, thereby enhancing the stability of the scene's light source. Moreover, we employ an image radiation intensity attenuation model to estimate the initial depth map. Finally, to refine the accuracy of depth estimation, we utilize a multi-scale residual fusion optimization technique., Results: To evaluate the performance of our proposed method, extensive experiments were conducted on public datasets. The structural similarity measures for continuous frames in three distinct clinical data scenes reached impressive values of 0.94, 0.82, and 0.84, respectively. These results demonstrate the effectiveness of our approach in capturing the intricate details of endoscopy images. Furthermore, the depth estimation accuracy achieved remarkable levels of 89.3 % and 91.2 % for the two models' data, respectively, underscoring the robustness of our method., Conclusions: Overall, the promising results obtained on public datasets highlight the significant potential of our method for clinical applications, facilitating reliable depth estimation and enhancing the quality of endoscopy surgical procedures., Competing Interests: Declaration of competing interest The authors declare that there are no conflicts of interest related to this article., (Copyright © 2023 Elsevier Ltd. All rights reserved.)

Fractional flow reserve (FFR) is considered as the gold standard for diagnosing coronary myocardial ischemia. Existing 3D computational fluid dynamics (CFD) methods attempt to predict FFR noninvasively using coronary computed tomography angiography (CTA). However, the accuracy and efficiency of the 3D CFD methods in coronary arteries are considerably limited. In this work, we introduce a multi-dimensional CFD framework that improves the accuracy of FFR prediction by estimating 0D patient-specific boundary conditions, and increases the efficiency by generating 3D initial conditions. The multi-dimensional CFD models contain the 3D vascular model for coronary simulation, the 1D vascular model for iterative optimization, and the 0D vascular model for boundary conditions expression. To improve the accuracy, we utilize clinical parameters to derive 0D patient-specific boundary conditions with an optimization algorithm. To improve the efficiency, we evaluate the convergence state using the 1D vascular model and obtain the convergence parameters to generate appropriate 3D initial conditions. The 0D patient-specific boundary conditions and the 3D initial conditions are used to predict FFR (FFR C ). We conducted a retrospective study involving 40 patients (61 diseased vessels) with invasive FFR and their corresponding CTA images. The results demonstrate that the FFR C and the invasive FFR have a strong linear correlation (r = 0.80, p < 0.001) and high consistency (mean difference: 0.014 ±0.071). After applying the cut-off value of FFR (0.8), the accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of FFR C were 88.5%, 93.3%, 83.9%, 84.8%, and 92.9%, respectively. Compared with the conventional zero initial conditions method, our method improves prediction efficiency by 71.3% per case. Therefore, our multi-dimensional CFD framework is capable of improving the accuracy and efficiency of FFR prediction significantly., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023 Elsevier Ltd. All rights reserved.)

Studies have found that different immune subtypes are present in the same tumor. Different tumor subtypes have different tumor microenvironments (TME). This means that the efficacy of immunotherapy in actual applications will, therefore, have different results. Existing tumor immune subtype studies have mostly focused on immune cells, stromal cells, genes and molecules without considering the presence of microbes. Some studies have shown that microflora can strongly promote many gastrointestinal cancers. The microbiome has, therefore, become an important biomarker and regulatory factor of cancer progression and therapeutic responses. In addition, the presence of microflora can strongly regulate the host immune system, indirectly affecting tumor growth. Taken together, it is important to study the relationships that develop among tumor tissue microorganisms, tumor immune subtype, and the TME. In this study, correlations between microbial abundance, immune cell infiltration, immune gene expression and tumor immune subtype were studied. To accomplish this, tissue microorganisms and immune cell ratios with significant differences between the different cancers were obtained by comparing 203 gastric cancer and intestinal cancer samples. Two immune subtypes of intestinal samples were obtained by K-means clustering algorithm and tissue microorganisms, immune cell ratios and immune-related genes with significant differences between different immune subtypes were screened through Wilcoxon rank sum test. The results showed that Clostridioides difficile, Aspergillus fumigatus, Yarrowia lipolytica, and Fusarium pseudograminearum were all closely associated with the identified tumor immune subtypes. Our open-source software is freely available from GitHub at https://github.com/gutmicrobes/IMM-subtype.git., Competing Interests: Declaration of competing interest None Declared., (Copyright © 2023. Published by Elsevier Ltd.)

Accurate detection of coronary artery stenosis in X-ray angiography (XRA) images is crucial for the diagnosis and treatment of coronary artery disease. However, stenosis detection remains a challenging task due to complicated vascular structures, poor imaging quality, and fickle lesions. While devoted to accurate stenosis detection, most methods are inefficient in the exploitation of spatio-temporal information of XRA sequences, leading to a limited performance on the task. To overcome the problem, we propose a new stenosis detection framework based on a Transformer-based module to aggregate proposal-level spatio-temporal features. In the module, proposal-shifted spatio-temporal tokenization (PSSTT) scheme is devised to gather spatio-temporal region-of-interest (RoI) features for obtaining visual tokens within a local window. Then, the Transformer-based feature aggregation (TFA) network takes the tokens as the inputs to enhance the RoI features by learning the long-range spatio-temporal context for final stenosis prediction. The effectiveness of our method was validated by conducting qualitative and quantitative experiments on 233 XRA sequences of coronary artery. Our method achieves a high F 1 score of 90.88%, outperforming other 15 state-of-the-art detection methods. It demonstrates that our method can perform accurate stenosis detection from XRA images due to the strong ability to aggregate spatio-temporal features., (Copyright © 2023 Elsevier Ltd. All rights reserved.)

Background: Marker-based augmented reality (AR) calibration methods for surgical navigation often require a second computed tomography scan of the patient, and their clinical application is limited due to high manufacturing costs and low accuracy., Methods: This work introduces a novel type of AR calibration framework that combines a Microsoft HoloLens device with a single camera registration module for surgical navigation. A camera is used to gather multi-view images of a patient for reconstruction in this framework. A shape feature matching-based search method is proposed to adjust the size of the reconstructed model. The double clustering-based 3D point cloud segmentation method and 3D line segment detection method are also proposed to extract the corner points of the image marker. The corner points are the registration data of the image marker. A feature triangulation iteration-based registration method is proposed to quickly and accurately calibrate the pose relationship between the image marker and the patient in the virtual and real space. The patient model after registration is wirelessly transmitted to the HoloLens device to display the AR scene., Results: The proposed approach was used to conduct accuracy verification experiments on the phantoms and volunteers, which were compared with six advanced AR calibration methods. The proposed method obtained average fusion errors of 0.70 ± 0.16 and 0.91 ± 0.13 mm in phantom and volunteer experiments, respectively. The fusion accuracy of the proposed method is the highest among all comparison methods. A volunteer liver puncture clinical simulation experiment was also conducted to show the clinical feasibility., Conclusions: Our experiments proved the effectiveness of the proposed AR calibration method, and revealed a considerable potential for improving surgical performance., (Copyright © 2022 Elsevier Ltd. All rights reserved.)

Background: Augmented reality (AR) based fusion imaging in endoscopic surgeries rely on the quality of image-to-patient registration and camera calibration, and these two offline steps are usually performed independently to get the target transformation separately. The optimal solution can be obtained under independent conditions but may not be globally optimal. All residual errors will be accumulated and eventually lead to inaccurate AR fusion., Methods: After a careful analysis of the principle of AR imaging, a robust online calibration framework was proposed for an endoscopic camera to enable accurate AR fusion. A 2D checkerboard-based homography estimation algorithm was proposed to estimate the local pose of the endoscopic camera, and the least square method was used to calculate the compensation matrix in combination with the optical tracking system., Results: In comparison with conventional methods, the proposed compensation method improved the performance of AR fusion, which reduced physical error by up to 82%, reduced pixel error by up to 83%, and improved target coverage by up to 6%. Experimental results of simulating mechanical noise revealed that the proposed compensation method effectively corrected the fusion errors caused by the rotation of the endoscopic tube without recalibrating the camera. Furthermore, the simulation results revealed the robustness of the proposed compensation method to noises., Conclusions: Overall, the experiment results proved the effectiveness of the proposed compensation method and online calibration framework, and revealed a considerable potential in clinical practice., (Copyright © 2021 Elsevier Ltd. All rights reserved.)