CADS: A Self-supervised Learner via Cross-modal Alignment and Deep Self-distillation for CT Volume Segmentation.

Self-supervised learning (SSL) has long had great success in advancing the field of annotation-efficient learning. However, when applied to CT volume segmentation, most SSL methods suffer from two limitations, including rarely using the information acquired by different imaging modalities and providing supervision only to the bottleneck encoder layer. To address both limitations, we design a pretext task to align the information in each 3D CT volume and the corresponding 2D generated X-ray image and extend self-distillation to deep self-distillation. Thus, we propose a self-supervised learner based on Cross-modal Alignment and Deep Self-distillation (CADS) to improve the encoder's ability to characterize CT volumes. The cross-modal alignment is a more challenging pretext task that forces the encoder to learn better image representation ability. Deep self-distillation provides supervision to not only the bottleneck layer but also shallow layers, thus boosting the abilities of both. Comparative experiments show that, during pre-training, our CADS has lower computational complexity and GPU memory cost than competing SSL methods. Based on the pre-trained encoder, we construct PVT-UNet for 3D CT volume segmentation. Our results on seven downstream tasks indicate that PVT-UNet outperforms state-of-the-art SSL methods like MOCOv3 and DiRA, as well as prevalent medical image segmentation methods like nnUNet and CoTr. Code and pre-trained weight will be available at https://github.com/yeerwen/CADS.

Structural insight into the subclass B1 metallo-ß-lactamase AFM-1.

The emergence of drug-resistant bacteria, facilitated by metallo-beta-lactamases (MBLs), presents a significant obstacle to the effective use of antibiotics in the management of clinical drug-resistant bacterial infections. AFM-1 is a MBL derived from Alcaligenes faecalis and shares 86% homology with the NDM-1 family. Both AFM-1 and NDM-1 demonstrate the ability to hydrolyze ampicillin and other ß-lactam antibiotics, however, their substrate affinities vary, and the specific reason for this variation remains unknown. We present the high-resolution structure of AFM-1. The active center of AFM-1 binds two zinc ions, and the conformation of the key amino acid residues in the active center is in accordance with that of NDM-1. However, the substrate-binding pocket of AFM-1 is considerably smaller than that of NDM-1. Additionally, the mutation of amino acid residues in the Loop3 region, as compared to NDM-1, results in the formation of a dense hydrophobic patch comprised of hydrophobic amino acid residues in this area, which facilitates substrate binding. Our findings lay the foundation for understanding the molecular mechanism of AFM-1 with a high affinity for substrates and provide a novel theoretical foundation for addressing the issue of drug resistance caused by B1 MBLs.

TriLA: Triple-Level Alignment Based Unsupervised Domain Adaptation for Joint Segmentation of Optic Disc and Optic Cup.

Cross-domain joint segmentation of optic disc and optic cup on fundus images is essential, yet challenging, for effective glaucoma screening. Although many unsupervised domain adaptation (UDA) methods have been proposed, these methods can hardly achieve complete domain alignment, leading to suboptimal performance. In this paper, we propose a triple-level alignment (TriLA) model to address this issue by aligning the source and target domains at the input level, feature level, and output level simultaneously. At the input level, a learnable Fourier domain adaptation (LFDA) module is developed to learn the cut-off frequency adaptively for frequency-domain translation. At the feature level, we disentangle the style and content features and align them in the corresponding feature spaces using consistency constraints. At the output level, we design a segmentation consistency constraint to emphasize the segmentation consistency across domains. The proposed model is trained on the RIGA+ dataset and widely evaluated on six different UDA scenarios. Our comprehensive results not only demonstrate that the proposed TriLA substantially outperforms other state-of-the-art UDA methods in joint segmentation of optic disc and optic cup, but also suggest the effectiveness of the triple-level alignment strategy.

Incidence, clearance, and factors associated with anal human papilloma virus infection among men with HIV.

OBJECTIVES: This study aims to estimate the incidence and persistence/clearance of anal human papilloma virus (HPV) infection and related factors among men with HIV in Taizhou, China. DESIGN: A prospective cohort study. METHODS: Men with HIV were recruited and followed up from 2016 to 2021. Questionnaire surveys were used to collect social-demographic and behavioral characteristics, and anal swabs were collected for HPV Genotyping. RESULTS: A total of 675 men with HIV were recruited and followed up. After an average follow-up time of 1.75 years, HPV39 (3.8/100 person-years), HPV52 (3.6/100 person-years), HPV51 (3.1/100 person-years), HPV58 (2.5/100 person-years) and HPV16 (2.4 cases/100 person-years) in the high-risk types showed the highest incidence rate. In marriage with woman [adjusted hazard ratio (aHR) = 0.44, 95% confidence interval (CI) 0.20-0.99] showed an inverse association with HPV incidence, while bisexuality or undetermined sexual orientation (aHR = 2.62, 95% CI 1.08-6.36) showed a positive association. For those infected at baseline, the top three high-risk HPV with the lowest clearance density were HPV52 (32.2/100 person-years), HPV58 (38.1/100 person-years), and HPV16 (43.5/100 person-years). Daily consumption of 1-28 g alcohol (aHR = 0.62, 95% CI 0.41-0.95) showed an inverse association with HPV clearance, while illicit drug use (aHR = 3.24, 95% CI 1.59-6.59) showed a positive association. CONCLUSION: Anal HPV infection and clearance were both active in men with HIV in China. Marriage status and sexuality were associated with the incidence of HPV infection, while substance use including alcohol and illicit drug were associated with HPV clearance. More studies are needed to explore the risk factors of HPV persistence.

ReFs: A hybrid pre-training paradigm for 3D medical image segmentation.

Self-supervised learning (SSL) has achieved remarkable progress in medical image segmentation. The application of an SSL algorithm often follows a two-stage training process: using unlabeled data to perform label-free representation learning and fine-tuning the pre-trained model on the downstream tasks. One issue of this paradigm is that the SSL step is unaware of the downstream task, which may lead to sub-optimal feature representation for a target task. In this paper, we propose a hybrid pre-training paradigm that is driven by both self-supervised and supervised objectives. To achieve this, a supervised reference task is involved in self-supervised learning, aiming to improve the representation quality. Specifically, we employ the off-the-shelf medical image segmentation task as reference, and encourage learning a representation that (1) incurs low prediction loss on both SSL and reference tasks and (2) leads to a similar gradient when updating the feature extractor from either task. In this way, the reference task pilots SSL in the direction beneficial for the downstream segmentation. To this end, we propose a simple but effective gradient matching method to optimize the model towards a consistent direction, thus improving the compatibility of both SSL and supervised reference tasks. We call this hybrid pre-training paradigm reference-guided self-supervised learning (ReFs), and perform it on a large-scale unlabeled dataset and an additional reference dataset. The experimental results demonstrate its effectiveness on seven downstream medical image segmentation benchmarks.

MFI-Net: Multiscale Feature Interaction Network for Retinal Vessel Segmentation.

Segmentation of retinal vessels on fundus images plays a critical role in the diagnosis of micro-vascular and ophthalmological diseases. Although being extensively studied, this task remains challenging due to many factors including the highly variable vessel width and poor vessel-background contrast. In this paper, we propose a multiscale feature interaction network (MFI-Net) for retinal vessel segmentation, which is a U-shaped convolutional neural network equipped with the pyramid squeeze-and-excitation (PSE) module, coarse-to-fine (C2F) module, deep supervision, and feature fusion. We extend the SE operator to multiscale features, resulting in the PSE module, which uses the channel attention learned at multiple scales to enhance multiscale features and enables the network to handle the vessels with variable width. We further design the C2F module to generate and re-process the residual feature maps, aiming to preserve more vessel details during the decoding process. The proposed MFI-Net has been evaluated against several public models on the DRIVE, STARE, CHASE_DB1, and HRF datasets. Our results suggest that both PSE and C2F modules are effective in improving the accuracy of MFI-Net, and also indicate that our model has superior segmentation performance and generalization ability over existing models on four public datasets.

A PrEP Demonstration Project Using eHealth and Community Outreach to Justice-Involved Cisgender Women and Their Risk Networks.

Women involved in criminal justice systems (WICJ) are a key population at risk for HIV, and pre-exposure prophylaxis (PrEP) is critical for HIV prevention. This project was designed to evaluate the feasibility and acceptability of delivering PrEP via eHealth to WICJ and members of their risk network (RN). We recruited HIV-negative cisgender WICJ index participants (n = 38) and risk network (RN) members (n = 67) using modified respondent-driven sampling. TDF/FTC was initiated for PrEP in participants meeting clinical criteria and dispensed through eHealth using a community-based, low barrier-to-care outreach model. Key steps in the PrEP care continuum were measured over 12 months. Enrolled participants (n = 105) had high current and lifetime justice-involvement and were predominantly cisgender women and non-Hispanic white with a mean age of 40.9y (SD 9.6). Despite most having primary care providers and medical insurance, PrEP awareness was low, and participants experienced high levels of medical, psychiatric, substance use, social, and economic need. Fifty-two participants (50%) were PrEP-eligible, of whom 24 (46%) initiated PrEP. TDF/FTC was safe and well-tolerated throughout follow-up and 13 individuals chose to remain on PrEP following study conclusion. In this novel PrEP demonstration project for WICJ and RN members, despite high medical, psychiatric, and social comorbidity, PrEP was positively received and effectively delivered using a community outreach model via eHealth.Registered on clinicaltrials.gov under trial registration number NCT03293290.

Increased CD4+ T cell count is associated with lower anal human papillomavirus prevalence among HIV-positive male cohort in Taizhou, China: a cross-sectional study.

OBJECTIVES: This study aims to investigate the association between CD4+ T cell count and combined antiretroviral therapy (cART) with the prevalence of anal human papillomavirus (HPV) infection among HIV-positive male cohort in China. METHODS: A survey was conducted in men from a HIV cohort in Taizhou, China between 2016 and 2019. A face-to-face questionnaire interview was administered, and an anal-canal swab was collected for HPV genotyping. RESULTS: A total of 766 HIV-positive men were recruited. The HPV prevalence was lower among those with increased CD4+ T cell count than those with decreased or unchanged (46.5 vs. 56.6%, p = 0.033) from baseline. In multivariable models, having the current CD4+ T cell count of 350-499 cells/µL (aOR 0.28, 95% CI 0.13-0.64), and of ≥ 500 cells/µL (aOR 0.26, 95% CI 0.11-0.60) were associated with lower prevalence of any type HPV infection compared with those with < 200 cells/µL. Having taken NVP + 3TC + AZT was inversely associated with any high-risk (HR)-HPV (aOR 0.47, 95% CI 0.25-0.90) and any low-risk (LR)-HPV infection (aOR 0.40, 95% CI 0.18-0.88), compared with those taking EFV + 3TC + TDF. CONCLUSIONS: Increased CD4+ T cell count at follow-up was significantly associated with lower prevalence of anal HPV infection. Inverse associations between NVP + 3TC + AZT and HR-HPV or LR-HPV infecton were observed.

Quantification of prostaglandins E2 and D2 using liquid chromatography-tandem mass spectrometry in a mouse ear edema model.

A sensitive, specific, and accurate high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) method was developed and validated for the quantification of prostaglandins D2 (PGD2) and E2 (PGE2) in a mouse ear edema model. We used activated charcoal to obtain PG-free ear samples. The chromatographic separation was performed using a Hypersil Gold C18 column. The limit of detection of each PG was 0.4 ng mL-1, and the intra- and inter-assay estimates of precision and accuracy were <14.5 and 94.2-102.9%, respectively. Stability studies showed that all analytes were stable under various storage conditions and analytical processes. The developed and validated method was successfully used to investigate the anti-inflammatory effects of cultured bear bile powder (CBBP) by quantitatively determining PGE2 and PGD2 levels in mouse ear edema samples. These results showed that CBBP significantly inhibited the xylene-induced ear edema in mice and reversed the xylene-induced elevation of PGE2 and PGD2 levels. These results provide useful data about the anti-inflammatory bioactivities in tissues, mediated by the reduction of PGE2 and PGD2 levels, and may further encourage research and development studies of CBBP for its use as an anti-inflammatory agent.

Study on the Relationships between Microscopic Cross-Linked Network Structure and Properties of Cyanate Ester Self-Reinforced Composites.

Bisphenol A dicyanate (BADCy) resin microparticles were prepared by precipitation polymerization synthesis and were homogeneously dispersed in a BADCy prepolymer matrix to prepare a BADCy self-reinforced composites. The active functional groups of the BADCy resin microparticles were characterized by Fourier transform infrared (FT-IR) spectroscopy. The results of an FT-IR curve showed that the BADCy resin microparticles had a triazine ring functional group and also had an active reactive group -OCN, which can initiate a reaction with the matrix. The structure of the BADCy resin microparticles was characterized by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). From the TEM results, the BADCy resin microparticles dispersed in the solvent were nano-sized and distributed at 40-60 nm. However, from the SEM results, agglomeration occurred after drying, the BADCy resin particels were micron-sized and distributed between 0.3 µm and 0.6 µm. The BADCy resin prepolymer was synthesized in our laboratory. A BADCy self-reinforced composite was prepared by using BADCy resin microparticles as a reinforcement phase. This corresponds to a composite in which the matrix and reinforcement phase are made from different morphologies of the same monomer. The DSC curve showed that the heat flow of the microparticles is different from the matrix during the curing reaction, this means the cured materials should be a microscopic two-phase structure. The added BADCy resin microparticles as reaction sites induced the formation of a more complete and regular cured polymer structure, optimizing the cross-linked network as well as increasing the interplay between the BADCy resin microparticles and prepolymer matrix. Relative to the neat BADCy resin material, the tensile strength, flexural strength, compressive strength and impact strength increased by 98.1%, 40.2%, 27.4%, and 85.4%, respectively. A particle toughening mechanism can be used to explain the improvement of toughness. The reduction in the dielectric constant showed that the cross-linked network of the self-reinforced composite was more symmetrical and less polar than the neat resin material, which supports the enhanced mechanical properties of the self-reinforced composite. In addition, the thermal behavior of the self-reinforced composite was characterized by thermogravimetric analysis (TGA) and dynamic mechanical thermal analysis (DMTA). The results of DMTA also establishes a basis for enhancing mechanical properties of the self-reinforced composite.

A further insight into the adsorption mechanism of protein on hydroxyapatite by FTIR-ATR spectrometry.

The adsorption mechanism of bovine serum albumin (BSA) on hydroxyapatite (HA) for different time intervals has been studied by Fourier transform infrared (FTIR)-attenuated total internal reflectance (ATR) spectrometry in this paper. The difference spectra obtained in HA and BSA frequency regions demonstrate that the binding of PO, from the phosphate (PO43-) of HA, to the hydrogen of methyl (-CH3), methene (-CH2) and amideII (-CNH) in the protein appears to be much faster and stronger than that of the PO group. In addition, Ca2+ must serve as a key role in the interaction of BSA with HA. The binding of Ca2+ to the oxygen of the peptide bond seems to induce a significant reconformation of polypeptide backbones from ß-pleated sheet to α-helix and ß-turn of helical circles. This alteration seems to have been accompanied by much hydrogen of polypeptides driven to bind PO43- and OH- of the HA actively and much -C=O and HN groups of the peptide bond freed from inter-chain hydrogen bonding to react on Ca2+ and combine strongly with the HA surface. This might be well expected to promote the HA biomineralization.

A further insight into the biosorption mechanism of Au(III) by infrared spectrometry.

BACKGROUND: The interactions of microbes with metal ions form an important basis for our study of biotechnological applications. Despite the recent progress in studying some properties of Au(III) adsorption and reduction by Bacillus megatherium D01 biomass, there is still a need for additional data on the molecular mechanisms of biosorbents responsible for their interactions with Au(III) to have a further insight and to make a better exposition. RESULTS: The biosorption mechanism of Au(III) onto the resting cell of Bacillus megatherium D01 biomass on a molecular level has been further studied here. The infrared (IR) spectroscopy on D01 biomass and that binding Au(III) demonstrates that the molecular recognition of and binding to Au(III) appear to occur mostly with oxygenous- and nitrogenous-active groups of polysaccharides and proteins in cell wall biopolymers, such as hydroxyl of saccharides, carboxylate anion of amino-acid residues (side-chains of polypeptide backbone), peptide bond (amide I and amide II bands), etc.; and that the active groups must serve as nucleation sites for Au(0) nuclei growth. A further investigation on the interactions of each of the soluble hydrolysates of D01, Bacillus licheniformis R08, Lactobacillus sp. strain A09 and waste Saccharomyces cerevisiae biomasses with Au(III) by IR spectrometry clearly reveals an essential biomacromolecule-characteristic that seems the binding of Au(III) to the oxygen of the peptide bond has caused a significant, molecular conformation-rearrangement in polypeptide backbones from ß-pleated sheet to α-helices and/or ß-turns of protein secondary structure; and that this changing appears to be accompanied by the occurrence, in the peptide bond, of much unbound -C=O and H-N- groups, being freed from the inter-molecular hydrogen-bonding of the ß-pleated sheet and carried on the helical forms, as well as by the alternation in side chain steric positions of protein primary structure. This might be reasonably expected to result in higher-affinity interactions of peptide bond and side chains with Au(III). CONCLUSIONS: The evidence suggests that the polypeptides appear to be activated by the intervention of Au(III) via the molecular reconformation and in turn react upon Au(III) actively and exert profound impacts on the course of Au(0) nucleation and crystal growth.

Alpha-synuclein functions as a negative regulator for expression of tyrosine hydroxylase.

Previous studies have shown that over-expression of alpha-Synuclein (alpha-Syn), a protein whose abnormality is implicated in the pathogenesis of Parkinson's disease (PD), reduces tyrosine hydroxylase (TH) expression and dopamine synthesis. To explore the possible mechanism for the regulation of TH expression by alpha-Syn, luciferase reporter gene carrying a -493/+27bp fragment of human TH gene (pGL3-TH520) and pcDNA carrying halpha-Syn gene (pcDNA-halpha-Syn) were co-transfected into T293 cells. The results showed that alpha-Syn was only detected in pcDNA-halpha-Syn-transfected cells but not in pcDNA vector control cells. In alpha-Syn-transfected cells, the luciferase activity was dramatically reduced compared with the vector control cells. These results suggest that alpha-Syn may function as a negative regulator for TH expression by affecting the activity of TH promoter.

A further insight into the biosorption mechanism of Pt(IV) by infrared spectrometry.

BACKGROUND: Platinum nanomaterial is one of the significant noble metal catalysts, and the interaction of platinum with microbe is one of the key factors in influencing the size and the distribution of the platinum nanoparticles on the microbial biomass. Some properties of Pt(IV) adsorption and reduction by resting cells of Bacillus megatherium D01 biomass have once been investigated, still the mechanism active in the platinum biosorption remains to be seen and requires further elucidating. RESULT: A further insight into the biosorption mechanism of Pt(IV) onto resting cells of Bacillus megatherium D02 biomass on a molecular level has been obtained. The image of scanning electron microscopy (SEM) of the D02 biomass challenged with Pt(IV) displayed a clear distribution of bioreduced platinum particles with sizes of nanometer scale on the biomass. The state of Pt(IV) bioreduced to elemental Pt(0) examined via X-ray photoelectron spectroscopy (XPS) suggested that the biomass reduces the Pt(IV) to Pt(II) followed by a slower reduction to Pt(0). The analysis of glucose content in the hydrolysates of D02 biomass for different time intervals using ultraviolet-visible (UV-vis) spectrophotometry indicated that certain reducing sugars occur in the hydrolyzed biomass and that the hydrolysis of polysaccharides of the biomass is a rapid process. The infrared (IR) spectrometry on D02 biomass and that challenged with Pt(IV), and on glucose and that reacted with Pt(IV) demonstrated that the interaction of the biomass with Pt(IV) seems to be through oxygenous or nitrogenous chemical functional groups on the cell wall biopolymers; that the potential binding sites for Pt species include hydroxyl of saccharides, carboxylate anion and carboxyl of amino acid residues, peptide bond, etc.; and that the free monosaccharic group bearing hemiacetalic hydroxyl from the hydrolyzed biomass behaving as an electron donor, in situ reduces the Pt(IV) to Pt(0). And moreover, the binding of the Pt(IV) to the oxygen of the carbonyl group of peptide bond caused a change in the secondary structure of proteins; i.e. a transformation, in polypeptide chains, of beta-folded to alpha-helical form; it might be expected to be more advantageous than beta-folded form to the platinum nanoparticles under shelter from gathering although the both special conformations of proteins could be much probably responsible for the stabilization of the particles. CONCLUSION: That knowledge could serve as a guide in the researches for improving the preparation of highly dispersive supported platinum catalyst and for fabricating new advanced platinum nanostructured devices by biotechnological methods.