Reasoning and Tuning: Graph Attention Network for Occluded Person Re-identification

Occluded person re-identification (re-id) aims to match occluded person images to holistic ones. Most existing works focus on matching collective-visible body parts by discarding the occluded parts. However, only preserving the collective-visible body parts causes great semantic loss for occluded images, decreasing the confidence of feature matching. On the other hand, we observe that the holistic images can provide the missing semantic information for occluded images of the same identity. Thus, compensating the occluded image with its holistic counterpart has the potential for alleviating the above limitation. In this paper, we propose a novel Reasoning and Tuning Graph Attention Network (RTGAT), which learns complete person representations of occluded images by jointly reasoning the visibility of body parts and compensating the occluded parts for the semantic loss. Specifically, we self-mine the semantic correlation between part features and the global feature to reason the visibility scores of body parts. Then we introduce the visibility scores as the graph attention, which guides Graph Convolutional Network (GCN) to fuzzily suppress the noise of occluded part features and propagate the missing semantic information from the holistic image to the occluded image. We finally learn complete person representations of occluded images for effective feature matching. Experimental results on occluded benchmarks demonstrate the superiority of our method.

Lyophilized mRNA-lipid nanoparticle vaccines with long-term stability and high antigenicity against SARS-CoV-2

Advanced mRNA vaccines play vital roles against SARS-CoV-2. However, most current mRNA delivery platforms need to be stored at -20 °C or -70 °C due to their poor stability, which severely restricts their availability. Herein, we develop a lyophilization technique to prepare SARS-CoV-2 mRNA-lipid nanoparticle vaccines with long-term thermostability. The physiochemical properties and bioactivities of lyophilized vaccines showed no change at 25 °C over 6 months, and the lyophilized SARS-CoV-2 mRNA vaccines could elicit potent humoral and cellular immunity whether in mice, rabbits, or rhesus macaques. Furthermore, in the human trial, administration of lyophilized Omicron mRNA vaccine as a booster shot also engendered strong immunity without severe adverse events, where the titers of neutralizing antibodies against Omicron BA.1/BA.2/BA.4 were increased by at least 253-fold after a booster shot following two doses of the commercial inactivated vaccine, CoronaVac. This lyophilization platform overcomes the instability of mRNA vaccines without affecting their bioactivity and significantly improves their accessibility, particularly in remote regions.

Production of tailored hydroxylated prodiginine showing combinatorial activity with rhamnolipids against plant-parasitic nematodes

Bacterial secondary metabolites exhibit diverse remarkable bioactivities and are thus the subject of study for different applications. Recently, the individual effectiveness of tripyrrolic prodiginines and rhamnolipids against the plant-parasitic nematode Heterodera schachtii, which causes tremendous losses in crop plants, was described. Notably, rhamnolipid production in engineered Pseudomonas putida strains has already reached industrial implementation. However, the non-natural hydroxyl-decorated prodiginines, which are of particular interest in this study due to a previously described particularly good plant compatibility and low toxicity, are not as readily accessible. In the present study, a new effective hybrid synthetic route was established. This included the engineering of a novel P. putida strain to provide enhanced levels of a bipyrrole precursor and an optimization of mutasynthesis, i.e., the conversion of chemically synthesized and supplemented monopyrroles to tripyrrolic compounds. Subsequent semisynthesis provided the hydroxylated prodiginine. The prodiginines caused reduced infectiousness of H. schachtii for Arabidopsis thaliana plants resulting from impaired motility and stylet thrusting, providing the first insights on the mode of action in this context. Furthermore, the combined application with rhamnolipids was assessed for the first time and found to be more effective against nematode parasitism than the individual compounds. To obtain, for instance, 50% nematode control, it was sufficient to apply 7.8 μM hydroxylated prodiginine together with 0.7 μg/ml (~ 1.1 μM) di-rhamnolipids, which corresponded to ca. ¼ of the individual EC₅₀ values. In summary, a hybrid synthetic route toward a hydroxylated prodiginine was established and its effects and combinatorial activity with rhamnolipids on plant-parasitic nematode H. schachtii are presented, demonstrating potential application as antinematodal agents. Graphical Abstract.

The substrate specificity in the O-demethylation of 4-alkylguaiacols by Cytochrome P450 AgcAP450

Alkylguaiacols are lignin-derived products obtained by reductive catalytic fractionation (RCF) of lignocellulosic biomass. Recently discovered AgcAP450 enzyme from CYP255A1 family oxidizes a range of 4-alkylguaiacols, having a preference for bulkier...

Research of Low-Temperature Performance of Polyphosphoric Acid-Modified Asphalt

Polyphosphoric acid (PPA) modifier, which can effectively improve the rheological properties of asphalt, is widely used in pavement engineering. In order to accurately evaluate the low-temperature performance of PPA-modified asphalt, in this study, PPA-modified asphalt and PPA/SBR-modified asphalt were prepared. The modification mechanism was explored by scanning electron microscopy (SEM) and fourier transform infrared spectroscopy (FTIR). Bending Beam Rheology (BBR) test was carried out, and four indexes, including K index, viscous flow (η₁), low-temperature integrated flexibility (Jc), and relaxation time (λ), were obtained by combining the Burgers model. The optimal low-temperature performance evaluation index of modified asphalt was determined by the analytic hierarchy process (AHP). The test results show that PPA addition to asphalt will produce chemical reactions, which can effectively improve the compatibility between SBR and neat asphalt. In the multi-index evaluation based on K, η₁, Jc, and λ, the same optimum content of PPA was obtained. AHP analysis further demonstrates that Jc is the optimal evaluation index for laboratory research on the low-temperature performance of PPA-modified asphalt, and λ index is the ideal evaluation index for the low-temperature performance of asphalt in engineering applications.

Is metformin a possible treatment for diabetic neuropathy?

Metformin is a hypoglycemic drug widely used in the treatment of type 2 diabetes. It has been proven to have analgesic and neuroprotective effects. Metformin can reverse pain in rodents, such as diabetic neuropathic pain, neuropathic pain caused by chemotherapy drugs, inflammatory pain and pain caused by surgical incision. In clinical use, however, metformin is associated with reduced plasma vitamin B12 levels, which can further neuropathy. In rodent diabetes models, metformin plays a neuroprotective and analgesic role by activating adenosine monophosphate-activated protein kinase, clearing methylgloxal, reducing insulin resistance, and neuroinflammation. This paper also summarized the neurological adverse reactions of metformin in diabetic patients. In addition, whether metformin has sexual dimorphism needs further study.

Prior Knowledge-Aware Fusion Network for Prediction of Macrovascular Invasion in Hepatocellular Carcinoma

Macrovascular invasion (MaVI) is a major threat to survival in hepatocellular carcinoma (HCC), which should be treated as early as possible to ensure safety and efficacy. In this aspect, MaVI prediction can be helpful. However, MaVI prediction is difficult because of the inter-class similarity and intra-class variation of HCC in computed tomography (CT) images. Moreover, existing methods fail to include clinical priori knowledge associated with HCC, leading to incomprehensive information extraction. In this paper, we proposed a prior knowledge-aware fusion network (PKAFnet) to accurately achieve MaVI prediction in CT images. First, a perception module was presented to extract features related to tumor marginal heterogeneity in the graph domain, which contributed to rotation invariance and captured intensity variations of tumor margin. Second, a tumor segmentation network was built to obtain global information of a 3D tumor image and information associated with tumor internal heterogeneity in the image domain. Finally, multi-domain features associated with the tumor margin and tumor region were combined by using a multi-domain attentional feature fusion module. Thus, by incorporating MaVI-related prior knowledge, our PKAFnet can alleviate overfitting, which can improve the discriminative ability. The proposed PKAFnet was validated on a multi-center dataset, and remarkable performance was achieved in an independent testing set. Moreover, the interpretability of perception module and segmentation network were presented in our paper, which illustrated the effectiveness and credibility of PKAFnet. Therefore, the proposed method showed great application potential for MaVI prediction.

[Development of a risk assessment scale and test of its validity and reliability for venous thromboembolism in adult burn patients]

Objective: To develop a venous thromboembolism (VTE) risk assessment scale for adult burn patients and to test its reliability and validity. Methods: The scale research method and multi-center cross-sectional survey method were used. Based on the results of literature analysis method and brain-storming method, the letter questionnaire for experts was formulated. Then 27 experts (9 doctors of burn department, 9 vascular surgeons, and 9 nurses) were performed with two rounds of correspondences by Delphi method, and the reliability of the experts was analyzed. The weight of each item was determined by optimal sequence diagram method and expert importance evaluation to form the VTE Risk Assessment Scale for Adult Burn Patients. A total of 223 adult burn inpatients, who were admitted to 5 tier Ⅲ grade A general hospitals including the Affiliated Hospital of Southwest Medical University, West China Hospital of Sichuan University, the Affiliated Hospital of North Sichuan Medical College, Nanchong Central Hospital, and the Second People's Hospital of Yibin City from October 1st 2019 to January 1st 2020, were selected as respondents by convenience sampling method. The first assessment was performed with the VTE Risk Assessment Scale for Adult Burn Patients within 24 hours of admission of patients, and real-time assessment was performed as the patients' condition and treatment changed. The highest value was taken as the result. Correlation coefficient method and critical ratio method were used for item analysis; Cronbach's α coefficient was used to test the internal consistency of scale; content validity index was used to analyze the content validity of the scale, and receiver's operating characteristic (ROC) curve was drawn to test the predictive validity of the scale. Data were statistically analyzed with chi-square test, Pearson correlation analysis, independent sample t test, and Z test. Results: As four questionnaires in the first round of correspondence were rejected as unqualified, and another 4 experts were selected for the 2 rounds of correspondence. Most of them were aged 41 to 50 years with postgraduate degrees, engaging in the current profession for 11 to 30 years, and all of them had professional titles of associate senior or above. The scale, constructed through literature analysis, group brainstorming, and two rounds of correspondence, includes 3 primary items and 50 secondary items. In the first round of correspondence, the recovery rate of valid questionnaires and the ratio with expert opinions were 85.2% (23/27) and 47.8% (11/23), respectively. In the second round of correspondence, the recovery rate of valid questionnaires and the ratio with expert opinions were 100% (27/27) and 11.1% (3/27), respectively. The average collective authority coefficients of experts were both 0.90 in the 2 rounds of correspondence. The mean values of importance assignment, full score rate, and selection rate above 4 were 4.21, 52.5%, and 77.2%, respectively, in the first round of correspondence, and 4.28, 45.2%, and 85.8%, respectively, in the second round of correspondence. The mean coefficients of variation and the mean value of Kendall's coefficient of harmony for each item were 0.21 and 0.30 in the first round of correspondence, respectively, and 0.16 and 0.36 in the second round of correspondence, respectively. In the first and second rounds of correspondence, the Kendall's coefficients of harmony of 3 primary items (age and underlying diseases, burn injury factors, and burn treatment factors) and total secondary items were statistically significant (with χ2 values of 121.46, 107.09, 116.00, 331.97, 169.97, 152.12, 141.54, and 471.70, P<0.01). The weights of primary items for age and underlying diseases, burn injury factors, and burn treatment factors were 0.04, 0.05, and 0.07, respectively. The weights of secondary items ranged from 0.71 to 0.99, with assigned values of 3 to 6. The total burn area of 223 patients ranged from 1% to 89% total body surface area, and the patients were aged from 19 to 96 years, with the risk assessment score from 0 to 98. Nine patients developed VTE, with a risk assessment score of 41 to 90. The scores of 37 items were significantly positively correlated with the total score of scale (with r values of 0.14 to 0.61, P<0.05 or P<0.01), and the items were retained. There were 36 secondary items with statistically significant differences between the patients in high-score group and low-score group (with Z values of -4.88 to -2.09, t values of -11.63 to -2.09, P<0.05 or P<0.01), and the items were retained. The total Cronbach's α coefficient of scale was 0.88. The total content validity index of scale was 0.95. The optimal threshold of the scale for the diagnosis of VTE was 40, at which the sensitivity was 88.9%, the specificity was 87.4%, the Youden index was 0.87, and the area under the ROC curve was 0.96 (with 95% confidence interval of 0.93 to 0.99, P<0.01). Conclusions: The age and underlying diseases, burn injury factors, and burn treatment factors are the risk factors for VTE in adult burn patients. The VTE risk assessment scale for adult burn patients developed based on these factors has good reliability and validity, and provide good reference value for clinical VTE risk assessment.

[Efficacy and safety of omalizumab in patients with refractory allergic asthma: a meta-analysis]

Objective: To evaluate the efficacy and safety of omalizumab in patients with refractory allergic asthma using meta-analysis. Methods: We searched databases including PubMed, Web of Science, Embase and the website of ClinicalTrials. gov registry for randomized controlled trials (RCTs), using the search terms: ("anti-IgE" OR "anti-immunoglobulin E" OR "anti-IgE antibody" OR "omalizumab" OR "rhuMAb-E25" OR "Xolair") AND ("allergic asthma"). The time was up to September 19th 2020. Review Manager 5.4 software and Stata16 software were used to calculate pooled RR or WMD, perform heterogeneity test, and assess publication bias. Results: Fifteen RCTs with 6 316 patients in total (omalizumab, n=3 469; placebo, n=2 847) met our selection criteria. Comparing with placebo, omalizumab reduced the risk of asthma exacerbations during both stable-inhaled corticosteroid (ICS) phase (RR=0.69, 95%CI: 0.63-0.75, P<0.001; I²=39.0%, P=0.090) and ICS-reduction phase (RR=0.55, 95%CI: 0.46-0.66, P<0.001; I²=41.0%, P=0.180), reduced emergency visits (RR=0.53, 95%CI: 0.38-0.73, P<0.001; I²=0, P=0.420), made a significant reduction in dosage of ICS (RR=1.35, 95%CI: 1.25-1.45, P<0.001; I²=22.0%, P=0.280) and even withdrew from ICS completely (RR=1.80, 95%CI: 1.41-2.31, P<0.001; I²=57.0%, P=0.070). Omalizumab significantly improved asthma-related quality of life (RR=1.81, 95%CI: 1.51-2.17). The use of rescue bronchodilators was significantly reduced in the omalizumab group (RR=0.78, 95%CI: 0.67-0.92) but there was no significant difference in the dosage of rescue bronchodilators (puff per day) (WMD=-0.32, 95%CI: -0.77-0.13). Patients taking omalizumab did not increase the frequency of any adverse events (RR=1.01, 95%CI: 0.98-1.03) and serious adverse events (RR=0.89, 95%CI: 0.74-1.06). Conclusions: Omalizumab is an ideal adjunctive treatment for refractory allergic asthma with good efficacy and safety. Further randomized controlled trials are needed to determine the appropriate duration of treatment.

Joint-Channel-Connectivity-Based Feature Selection and Classification on fNIRS for Stress Detection in Decision-Making

Stress is one of the contributing factors affecting decision-making. Therefore, early stress recognition is essential to improve clinicians' decision-making performance. Functional near-infrared spectroscopy (fNIRS) has shown great potential in detecting stress. However, the majority of previous studies only used fNIRS features at the individual level for classification without considering the correlations among channels corresponding to the brain, which may provide distinguishing features. Hence, this study proposes a novel joint-channel-connectivity-based feature selection and classification algorithm for fNIRS to detect stress in decision-making. Specifically, this approach integrates feature selection and classifier modeling into a sparse model, where intra- and inter-channel regularizers are designed to explore potential correlations among channels to obtain discriminating features. In this paper, we simulated the decision-making of medical students under stress through the Trier Social Stress Test and the Balloon Analog Risk Task and recorded their cerebral hemodynamic alterations by fNIRS device. Experimental results illustrated that our method with the accuracy of 0.961 is superior to other machine learning methods. Additionally, the stress correlation and connectivity of brain regions calculated by feature selection have been confirmed in previous studies, which validates the effectiveness of our method and helps optimize the channel settings of fNIRS. This work was the first attempt to utilize a sparse model that simultaneously considers the sparsity of features and the correlation of brain regions for stress detection and obtained an admirable classification performance. Thus, the proposed model might be a useful tool for medical personnel to automatically detect stress in clinical decision-making situations.

[Application of anti-programmed cell death protein 1 antibody in children with refractory or relapsed Hodgkin lymphoma]

Objective: To explore the efficacy of anti-programmed cell death protein 1 (PD-1) antibody for children with refractory or relapsed Hodgkin lymphoma. Methods: Clinical data including short-term efficacy, long-term efficacy and adverse reaction of Hodgkin lymphoma children treated with anti-PD-1 antibody in Beijing Children's Hospital, Capital Medical University from December 2017 to June 2021 were analyzed retrospectively. Efficacy was evaluated as complete remission (CR), partial remission (PR), stable disease (SD), progressed disease (PD) and no response. Results: A total of 6 cases were included, including 5 males and 1 female. The age at the start of anti-PD-1 antibody treatment was 11.6 (10.2, 13.3) years, including 3 cases of mixed cellularity type, 2 cases of nodular sclerosis type, and 1 case of nodular lymphocyte-predominant type. There were 4 cases of stage Ⅳ and 2 cases of stage Ⅲ. All cases were assigned to the high-risk group, and 5 cases had B symptoms, all cases were refractory or relapsed Hodgkin lymphoma before the start of anti-PD-1 antibody treatment. Early evaluation showed that within 12 weeks of the treatment, 3 cases achieved PR and 3 cases achieved SD, while late evaluation showed that after 16 weeks of the treatment, 5 cases achieved CR and 1 case achieved PR. None of the children progressed during treatment. The follow-up time was 27 (21,41) months. Among all cases, 5 cases had event-free survival and sustained remission, 1 case had fever about 4 weeks after the drug withdrawal, finally he was confirmed to be transformed to B cell lymphoma between diffuse B cells and classic Hodgkin lymphoma. All the patients were well tolerated with anti-PD-1 antibody therapy. No adverse reactions such as high fever, chills, rash, etc. were observed during infusion. None of the patients were delayed, dose reduction or withdrawal due to adverse reactions. Two cases had reactive skin vascular hyperplasia during the treatment, without pain or itching, and they recovered on their own after stopping anti-PD-1 antibody therapy without other special treatment. Conclusion: Anti-PD-1 monoclonal antibody for children with refractory or relapsed Hodgkin lymphoma have good efficacy and tolerable side effects.

2D-3D cascade network for glioma segmentation in multisequence MRI images using multiscale information

BACKGROUND AND OBJECTIVE

Glioma segmentation is an important procedure for the treatment plan and follow-up evaluation of patients with glioma. UNet-based networks are widely used in medical image segmentation tasks and have achieved state-of-the-art performance. However, context information along the third dimension is ignored in 2D convolutions, whereas difference between z-axis and in-plane resolutions is large in 3D convolutions. Moreover, an original UNet structure cannot capture fine details because of the reduced resolution of feature maps near bottleneck layers.

METHODS

To address these issues, a novel 2D-3D cascade network with multiscale information module is proposed for the multiclass segmentation of gliomas in multisequence MRI images. First, a 2D network is applied to fully exploit potential intra-slice features. A variational autoencoder module is incorporated into 2D DenseUNet to regularize a shared encoder, extract useful information, and represent glioma heterogeneity. Second, we integrated 3D DenseUNet with the 2D network in cascade mode to extract useful inter-slice features and alleviate the influence of large difference between z-axis and in-plane resolutions. Moreover, a multiscale information module is used in the 2D and 3D networks to further capture the fine details of gliomas. Finally, the whole 2D-3D cascade network is trained in an end-to-end manner, where the intra-slice and inter-slice features are fused and optimized jointly to take full advantage of 3D image information.

RESULTS

Our method is evaluated on publicly available and clinical datasets and achieves competitive performance in these two datasets.

CONCLUSIONS

These results indicate that the proposed method may be a useful tool for glioma segmentation.

POS0482 LONG NON-CODING RNA H19X IS A MEDIATOR OF ENDOTHELIAL CELL ACTIVATION IN SYSTEMIC SCLEROSIS

Background

In one of our previous studies, we demonstrated that long non-coding RNA (lncRNA) H19X plays a crucial role in the development of TGFβ driven fibrosis in systemic sclerosis (SSc) and other fibrotic diseases1.

Objectives

To define the functional relevance of H19X in endothelial cell (EC) activation as a decisive process in SSc vasculopathy.

Methods

Correlation of H19X expression and microvascular gene signature was computed on bulk RNA-Seq data derived from SSc skin biopsies of patients enrolled in the multicentre Prospective Registry of Early Systemic Sclerosis cohort (PRESS, n=48 SSc vs. n=33 healthy controls, HCs). Single cell RNA sequencing (scRNA-seq) data were collected from 27 diffuse cutaneous SSc (dcSSc) and 10 HC skin biopsies. Single cells were barcoded and encapsulated in droplets using a 10X Genomics system. After cDNA synthesis, the libraries were prepared and sequenced using Illumina NovaSeq-500 platform. Seurat package in R (v.3.0) was used to perform data analysis. EC were identified by enrichment of EC markers CLDN5, VWF and PECAM1. One thousand five hundred eighty-three and 3398 EC were identified from HC and SSc patients, respectively. Cells were analysed for the expression of H19X and EC activation markers. Additionally, differential expression and pathway enrichment analysis between H19X expressing cells and H19X negative cells was carried out. The function of H19X was investigated in human dermal microvascular EC (HDMEC) by silencing, using locked nucleic acid antisense oligonucleotides (LNA GapmeRs). Gene expression was measured by qPCR. Protein levels of endothelial adhesion molecules were analysed by Western Blot. Endothelial adhesion was evaluated by co-culture of HDMEC and fluorescently labelled peripheral blood mononuclear cells (PBMCs).

Results

H19X expression was found significantly upregulated in SSc skin biopsies of the PRESS cohort (p<0.0001). The expression of H19X positively correlated with the microvascular endothelial cell gene signature in all subjects (SSc and HC, R=0.43, p<0.0001), confirming that H19X is expressed in this cell type. To determine if H19X might be an important factor in SSc EC dysfunction scRNAseq was performed. This analysis revealed a significant upregulation of H19X in SSc EC as compared to HC EC (p=0.0095). H19X was found to be upregulated in several EC subclusters including arterial (SEMA3G, HEY1), capillary (CA4, RGCC), venous (ACKR1, VCAM1) and lymphatic (PROX1, LYVE1). H19X displayed highest expression in injured SSc EC and capillary SSc EC. Co-expression analysis of the scRNA-seq data revealed higher expression of several adhesion molecules in EC expressing H19X, including VCAM1, ICAM and JAM3. KEGG pathway analysis revealed that differentially expressed genes in H19X expressing cells were highly associated with the ‘Cell adhesion molecule’ pathway (p=2.209e-7). H19X silencing lead to a significant downregulation of mRNA levels of genes encoding adhesion molecules VCAM1 (n=7, p<0.05) and E-selectin (n=7, p<0.01) at 48h after transfection. VCAM1, but not E-Selectin, was also reduced at protein level as revealed by Western Blot (n=3). The functional relevance of H19X on endothelial adhesion was confirmed by PBMCs with H19X silenced HDMEC where we were able to demonstrate a significant decrease in leucocyte-to-endothelial cell adhesion (n=5, p<0.05).

Conclusion

Our results show that lncRNA H19X could contribute to EC activation in SSc vasculopathy, acting as a regulator of expression of adhesion molecules in EC.

References

[1]Pachera E, et al. Long noncoding RNA H19X is a key mediator of TGF-β-driven fibrosis. J Clin Invest. 2020 Sep 1;130(9):4888-4905.

Disclosure of Interests

Francesca Tirelli: None declared, Elena Pachera: None declared, Robert Lafyatis Consultant of: Pfizer, Bristol Myers Squibb, Boehringer-Ingleheim, Formation, Sanofi, Boehringer-Mannheim, Merck and Genentech/Roche, Grant/research support from: Corbus, Formation, Moderna, Regeneron, Pfizer and Kiniksa, Menqi Huang: None declared, Shervin Assassi: None declared, Eva Camarillo: None declared, Francesco Zulian: None declared, Gabriela Kania: None declared, Oliver Distler Speakers bureau: Bayer, Boehringer Ingelheim, Janssen, Medscape, Consultant of: Abbvie, Acceleron, Alcimed, Amgen, AnaMar, Arxx, AstraZeneca, Baecon, Blade, Bayer, Boehringer Ingelheim, Corbus, CSL Behring, 4P Science, Galapagos, Glenmark, Horizon, Inventiva, Kymera, Lupin, Miltenyi Biotec, Mitsubishi Tanabe, MSD, Novartis, Prometheus, Roivant, Sanofi and Topadur, Grant/research support from: Kymera, Mitsubishi Tanabe, Boehringer Ingelheim.