Kang Liao

Purpose: Candidozyma auris (C. auris) is an emerging fungal pathogen that is resistant to multiple drugs and poses a serious threat to global health. This study aimed to investigate the genetic diversity and antifungal resistance profiles of C. auris in southern China.Methods: A total of 108 clinical C. auris isolates were collected from eight hospitals in Guangdong Province between January 2023 and January 2024. All the isolates were identified via matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS). Phylogenetic analysis was conducted on the basis of the sequencing results of the RPB1 and D1/D2 genes. The mechanisms of resistance to fluconazole and echinocandins were investigated through sequencing of the ERG11 and FKS1 genes. Additionally, whole-genome sequencing was performed on echinocandin-resistant and echinocandin-sensitive isolates to analyse genetic homology among the strains.Results: All the strains were classified into two genetic clades, clade I and clade III, and all exhibited resistance to fluconazole. In the fluconazole-resistant strains, the amino acid substitutions Y132F and VF125AL were identified in the ERG11 gene. The resistance rates to caspofungin, micafungin, and anidulafungin were 7.4, 7.4, and 3.7%, respectively. Among the eight echinocandin-resistant strains, amino acid substitutions (S639Y, W691L, and S639F) were found within the HS1 hotspot region of the FKS1 gene. A phylogenetic tree was constructed on the basis of 403 SNPs and revealed two major clusters: Cluster A and Cluster B. Cluster A included the 16 isolates analyzed in this study. Cluster B consisted of 12 reference isolates retrieved from publicly available genomic databases.Conclusion: Genetic clade I and clade III C. auris strains are prevalent in southern China and present high levels of resistance to fluconazole. Controlling the spread of C. auris in this region presents significant challenges for public health management.

The study aims were to present in vitro susceptibilities of clinical isolates from Gram-negative bacteria bloodstream infections (GNBSI) collected in China. GNBSI isolates were collected from 18 tertiary hospitals in 7 regions of China from 2018 to 2020. Minimum inhibitory concentrations were assessed using a Trek Diagnostic System. Susceptibility was determined using CLSI broth microdilution, and breakpoints were interpreted using CLSI M100 (2021). A total of 1,815 GNBSI strains were collected, with E. coli (42.4%) and Klebsiella pneumoniae (28.6%) being the most prevalent species, followed by P. aeruginosa (6.7%). Susceptibility analyses revealed low susceptibilities (E. coli and K. pneumonia to third-/fourth-generation cephalosporins, monobactamases, and fluoroquinolones. High susceptibilities to colistin (95.0%) and amikacin (81.3%) were found for K. pneumoniae, while Acinetobacter baumannii exhibited a high susceptibility (99.2%) to colistin but a low susceptibility to other antimicrobials (K. pneumoniae and A. baumannii than isolates from non-ICUs (all P < 0.05). Carbapenem-resistant and ESBL-producing K. pneumoniae detection was different across regions (both P < 0.05). E. coli and K. pneumoniae were major contributors to GNBSI, while A. baumannii exhibited severe drug resistance in isolates obtained from ICU departments.

Background: Though droplet digital PCR (ddPCR) has emerged as a promising tool for early pathogen detection in bloodstream infections (BSIs), more studies are needed to support its clinical application widely due to different ddPCR platforms with discrepant diagnostic performance. Additionally, there is still a lack of clinical data to reveal the association between pathogen loads detected by ddPCR and corresponding BSIs.Methods: In this prospective study, 173 patients with suspected BSIs were enrolled. A multiplex ddPCR assay was used to detect 18 pathogens. The results of ddPCR testing were evaluated in comparison with blood cultures (BCs) and clinical diagnosis. Taking BC as the gold standard, receiver operating characteristic curve and Cohen’s kappa agreement were used to investigate whether the pathogen load could predict a corresponding culture-proven BSI for the top five microorganisms detected by ddPCR.Results: Of the 173 blood samples collected, BC and ddPCR were positive in 48 (27.7%) and 92 (53.2%) cases, respectively. Compared to BC, the aggregate sensitivity and specificity for ddPCR were 81.3% and 63.2%, respectively. After clinical adjudication, the sensitivity and specificity of ddPCR increased to 88.8% and 86.0%, respectively. There were 143 microorganisms detected by ddPCR. The DNA loads of these microorganisms ranged from 30.0 to 3.2×10⁵ copies/mL (median level: 158.0 copies/mL), 72.7% (104/143) of which were below 1,000 copies/mL. Further, statistical analysis showed the DNA loads of Escherichia coli (AUC: 0.954, 95% CI: 0.898-1.000, κ=0.731, cut-off values: 93.0 copies/mL) and Klebsiella pneumoniae (AUC: 0.994, 95% CI: 0.986-1.000, κ=0.834, cut-off values: 196.5 copies/mL) were excellent predictors for the corresponding BSIs. The DNA loads of Pseudomonas aeruginosa (AUC: 0.816, 95% CI: 0.560-1.000, κ=0.167), Acinetobacter baumannii (AUC: 0.728, 95% CI: 0.195-1.000), and Enterococcus spp. (AUC: 0.282, 95% CI: 0.000-0.778) had little predictive value for the corresponding culture-proven BSIs.Conclusion: Our results indicate that the multiplex ddPCR is a promising platform as a complementary add-on to conventional BC. The DNA loads of E. coli and K. pneumoniae present excellent predictive value for the corresponding BSIs. Further research is needed to explore the predictive potential of ddPCR for other microorganisms.

Introduction: Preservation fluid (PF) contaminations are common in conventional liver transplantation (CLT) and presumably originate from organ or PF exposures to the external environment in a non-strict sterile manner. Such exposures and PF contamination may be avoided in ischaemia-free liver transplantation (IFLT) because of the strict sterile surgical procedures. In this study, the authors evaluated the impact of IFLT on organ PF contamination. Methods: A post-hoc analysis using data from the first randomized controlled trial of IFLT was performed to compare the incidence, pathogenic spectrum of PF contamination, and incidence of early recipient infection between IFLT and CLT. Multivariable logistic regression was used to explore risk factors for PF contamination. Results: Of the 68 cases recruited in the trial, 64 were included in this post-hoc analysis. The incidence of culture-positive PF was 9.4% (3/32) in the IFLT group versus 78.1% (25/32) in the CLT group (PP=0.196). Multivariate analysis revealed that adopting IFLT is an independent protective factor for culture-positive PF. Conclusion: PF contamination is substantially decreased in IFLT, and IFLT application is an independent protective factor for PF contamination. Using rigorous sterile measures and effective antibiotic therapy during IFLT may decrease PF contamination.

The wide-angle lens shows appealing applications in VR technologies, but it introduces severe radial distortion into its captured image. To recover the realistic scene, previous works devote to rectifying the content of the wide-angle image. However, such a rectification solution inevitably distorts the image boundary, which changes related geometric distributions and misleads the current vision perception models. In this work, we explore constructing a win-win representation on both content and boundary by contributing a new learning model, i.e., Rectangling Rectification Network (RecRecNet). In particular, we propose a thin-plate spline (TPS) module to formulate the non-linear and non-rigid transformation for rectangling images. By learning the control points on the rectified image, our model can flexibly warp the source structure to the target domain and achieves an end-to-end unsupervised deformation. To relieve the complexity of structure approximation, we then inspire our RecRecNet to learn the gradual deformation rules with a DoF (Degree of Freedom)-based curriculum learning. By increasing the DoF in each curriculum stage, namely, from similarity transformation (4-DoF) to homography transformation (8-DoF), the network is capable of investigating more detailed deformations, offering fast convergence on the final rectangling task. Experiments show the superiority of our solution over the compared methods on both quantitative and qualitative evaluations. The code and dataset are available at https://github.com/KangLiao929/RecRecNet.

Although the distortion correction of fisheye images has been extensively studied, the correction of fisheye videos is still an elusive challenge. For different frames of the fisheye video, the existing image correction methods ignore the correlation of sequences, resulting in temporal jitter in the corrected video. To solve this problem, we propose a temporal weighting scheme to get a plausible global optical flow, which mitigates the jitter effect by progressively reducing the weight of frames. Subsequently, we observe that the inter-frame optical flow of the video is facilitated to perceive the local spatial deformation of the fisheye video. Therefore, we derive the spatial deformation through the flows of fisheye and distorted-free videos, thereby enhancing the local accuracy of the predicted result. However, the independent correction for each frame disrupts the temporal correlation. Due to the property of fisheye video, a distorted moving object may be able to find its distorted-free pattern at another moment. To this end, a temporal deformation aggregator is designed to reconstruct the deformation correlation between frames and provide a reliable global feature. Our method achieves an end-to-end correction and demonstrates superiority in correction quality and stability compared with the SOTA correction methods.

Not everybody can be equipped with professional photography skills and sufficient shooting time, and there can be some tilts in the captured images occasionally. In this paper, we propose a new and practical task, named Rotation Correction, to automatically correct the tilt with high content fidelity in the condition that the rotated angle is unknown. This task can be easily integrated into image editing applications, allowing users to correct the rotated images without any manual operations. To this end, we leverage a neural network to predict the optical flows that can warp the tilted images to be perceptually horizontal. Nevertheless, the pixel-wise optical flow estimation from a single image is severely unstable, especially in large-angle tilted images. To enhance its robustness, we propose a simple but effective prediction strategy to form a robust elastic warp. Particularly, we first regress the mesh deformation that can be transformed into robust initial optical flows. Then we estimate residual optical flows to facilitate our network the flexibility of pixel-wise deformation, further correcting the details of the tilted images. To establish an evaluation benchmark and train the learning framework, a comprehensive rotation correction dataset is presented with a large diversity in scenes and rotated angles. Extensive experiments demonstrate that even in the absence of the angle prior, our algorithm can outperform other state-of-the-art solutions requiring this prior. The code and dataset are available at https://github.com/nie-lang/RotationCorrection.

Although the distortion correction of fisheye images has been extensively studied, the correction of fisheye videos is still an elusive challenge. For different frames of the fisheye video, the existing image correction methods ignore the correlation of sequences, resulting in temporal jitter in the corrected video. To solve this problem, we propose a temporal weighting scheme to get a plausible global optical flow, which mitigates the jitter effect by progressively reducing the weight of frames. Subsequently, we observe that the inter-frame optical flow of the video is facilitated to perceive the local spatial deformation of the fisheye video. Therefore, we derive the spatial deformation through the flows of fisheye and distorted-free videos, thereby enhancing the local accuracy of the predicted result. However, the independent correction for each frame disrupts the temporal correlation. Due to the property of fisheye video, a distorted moving object may be able to find its distorted-free pattern at another moment. To this end, a temporal deformation aggregator is designed to reconstruct the deformation correlation between frames and provide a reliable global feature. Our method achieves an end-to-end correction and demonstrates superiority in correction quality and stability compared with the SOTA correction methods.

Fisheye image rectification has a long-term unresolved issue with synthetic-to-real generalization. In most previous works, the model trained on the synthetic images obtains unsatisfactory performance on the real-world fisheye image. To this end, we propose a Dual Diffusion Architecture (DDA) for the fisheye rectification with a better generalization ability. The proposed DDA is simultaneously trained with paired synthetic fisheye images and unlabeled real fisheye images. By gradually introducing noises, the synthetic and real fisheye images can eventually develop into a consistent noise distribution, improving the generalization and achieving unlabeled real fisheye correction. The original image serves as the prior guidance in existing DDPMs (Denoising Diffusion Probabilistic Models). However, the non-negligible indeterminate relationship between the prior condition and the target affects the generation performance. Especially in the rectification task, the radial distortion can cause significant artifacts. Therefore, we provide an unsupervised one-pass network that produces a plausible new condition to strengthen guidance. This network can be regarded as an alternate scheme for fast producing reliable results without iterative inference. Compared with the state-of-the-art methods, our approach can reach superior performance in both synthetic and real fisheye image corrections.

Background: The incidence of Clostridioides difficile infection (CDI) has increased in recent years in patients with inflammatory bowel disease (IBD). C. difficile is a toxin-producing bacterium, and CDI results in the worsening of underlying IBD, increasing the risk of IBD treatment failure, surgery, and hospitalization. Because the symptoms of CDI overlap with those of IBD, it is challenging to make a differential diagnosis. Therefore, early, rapid, and reliable diagnostic tools that can identify CDI in IBD patients would be valuable to clinicians. Methods: This study retrospectively collected 135 patients with IBD. Among them, 44 patients were diagnosed with CDI, and 42 patients were diagnosed with viral or fungal infections. A total of 49 patients without infections were defined as the control group. The diagnostic values of procalcitonin (PCT), C-reactive protein (CRP), and white blood cell (WBC) count in the peripheral blood were examined. Results: In this study, PCT levels were significantly higher in patients with CDI than in non-CDI patients (including patients with viral/fungal infections and the control group; p < 0.001 and p < 0.05, respectively). CRP levels were significantly higher in patients with CDI than in non-CDI patients (p < 0.05). The area under the curve (AUC) of PCT and WBC count were compared using DeLong’s test: the AUCs of PCT vs. CRP for the detection of the IBD–CDI group and the control group was 0.826 [95% confidence interval (CI) 0.743–0.909] vs. 0.663 [95% confidence interval (CI) 0.551–0.774] (p < 0.05), respectively. WBC count was inferior as a diagnostic tool for CDI. The sensitivity was 59.09% (95% CI: 43.2% to 73.7%), the specificity was 89.80% (95% CI: 77.8% to 96.6%), and the positive likelihood ratio LR (+) was 5.79 for PCT for the diagnosis of CDI. Conclusions: The present study demonstrates the superiority of PCT over CRP and WBC count for the rapid diagnosis of CDI in IBD patients.