PPDM++: Parallel Point Detection and Matching for Fast and Accurate HOI Detection.

Human-Object Interaction (HOI) detection aims to understand human activities by detecting interaction triplets. Previous HOI detection methods adopt a two-stage instance-driven paradigm. Unfortunately, many non-interactive human-object pairs generated by the first stage are the main obstacle impeding HOI detectors from high efficiency and promising performance. To remedy this, we propose a novel top-down interaction-driven paradigm, detecting interactions first and bridging interactive human-object pairs through interactions. We formulate HOI as a point triplet [Formula: see text]human point, interaction point, object point[Formula: see text] and design a Parallel Point Detection and Matching (PPDM) framework. We further take advantage of two-stage methods and propose a novel framework, PPDM++, that detects the interactive human-object pairs by PPDM, then extracts region features for each pair to predict actions. The core of PPDM/PPDM++ is to convert the instance-driven bottom-up paradigm to an interaction-driven top-down paradigm, thus avoiding additional computation costs from traversing a tremendous number of non-interactive pairs. Benefiting from the advanced paradigm, PPDM/PPDM++ has achieved significant performance gains with high efficiency. PPDM-DLA-34 has achieved 19.94 mAP with 42 FPS as the first real-time HOI detector, and PPDM++-SwinB achieves 30.1 mAP with 17 FPS on HICO-DET dataset. We also built an application-oriented database named HOI-A, a supplement to the existing datasets.

Delving Into the Devils of Bird's-Eye-View Perception: A Review, Evaluation and Recipe.

Learning powerful representations in bird's-eye-view (BEV) for perception tasks is trending and drawing extensive attention both from industry and academia. Conventional approaches for most autonomous driving algorithms perform detection, segmentation, tracking, etc., in a front or perspective view. As sensor configurations get more complex, integrating multi-source information from different sensors and representing features in a unified view come of vital importance. BEV perception inherits several advantages, as representing surrounding scenes in BEV is intuitive and fusion-friendly; and representing objects in BEV is most desirable for subsequent modules as in planning and/or control. The core problems for BEV perception lie in (a) how to reconstruct the lost 3D information via view transformation from perspective view to BEV; (b) how to acquire ground truth annotations in BEV grid; (c) how to formulate the pipeline to incorporate features from different sources and views; and (d) how to adapt and generalize algorithms as sensor configurations vary across different scenarios. In this survey, we review the most recent works on BEV perception and provide an in-depth analysis of different solutions. Moreover, several systematic designs of BEV approach from the industry are depicted as well. Furthermore, we introduce a full suite of practical guidebook to improve the performance of BEV perception tasks, including camera, LiDAR and fusion inputs. At last, we point out the future research directions in this area. We hope this report will shed some light on the community and encourage more research effort on BEV perception.

An anoikis-related lncRNA signature is a useful tool for predicting the prognosis of patients with lung adenocarcinoma.

Background: Anoikis-related long non-coding RNAs (ARLs) play a critical role in tumor metastasis and progression, suggesting that they may serve as risk markers for cancer. This study aimed to investigate the prognostic value of ARLs in patients with lung adenocarcinoma (LUAD). Methods: Clinical data, RNA sequencing (RNA-seq) data, and mutation data from the LUAD project were obtained from The Cancer Genome Atlas (TCGA) database. The Molecular Signatures Database (MSigDB) and the GeneCard database were used to collect an anoikis-related gene (ARG) set. Pearson correlation analysis was performed to identify ARLs. LASSO and Cox regression were then used to establish a prognostic risk signature for ARLs. The median risk score served as the basis for categorizing patients into high and low-risk groups. Kaplan-Meier analysis was utilized to compare the prognosis between these two groups. The study also examined the associations between risk scores and prognosis, clinicopathological characteristics, immune status, tumor mutation burden (TMB), and chemotherapeutic agents. LncRNA expression was assessed using quantitative real-time PCR (qRT-PCR). Results: A total of 480 RNA expression profiles, 501 ARGs, and 2698 ARLs were obtained from the database. A prognostic ARL signature for LUAD was established, consisting of 9 lncRNAs. Patients in the low-risk group exhibited significantly better prognosis compared to those in the high-risk group (P < 0.001). The 9 lncRNAs from the ARL signature were identified as independent prognostic factors (P < 0.001). The signature demonstrated high accuracy in predicting LUAD prognosis, with area under the curve values exceeding 0.7. The risk scores for ARLs showed strong negative correlations with stroma score (P = 5.9E-07, R = -0.23), immune score (P = 9.7E-09, R = -0.26), and microenvironment score (P = 8E-11, R = -0.29). Additionally, the low-risk group exhibited significantly higher TMB compared to the high-risk group (P = 4.6E-05). High-risk status was significantly associated with lower half-maximal inhibitory concentrations for most chemotherapeutic drugs. Conclusion: This newly constructed signature based on nine ARLs is a useful instrument for the risk stratification of LUAD patients. The signature has potential clinical significance for predicting the prognosis of LUAD patients and guiding personalized immunotherapy.

A New perspective: How Pathogens Manipulate Phagocytosis?

The immune system is a complicated, closely regulated mechanism that evolved to keep people healthy from infectious pathogens. Phagocytosis is important for both innate and acquired immunity, which is a critical process for microbial pathogens and apoptotic cells to be consumed and eliminated. However, several pathogens have evolved different strategies to escape detection and killing by phagocytosis. Recently, with the increase in infectious diseases and antibiotic resistance, it is significant for people to have a deep understanding of immune evasion, which may become an opportunity to explore new treatments and vaccination. Additionally, researchers mostly study immune evasion of a single pathogen but rarely summarize pathogens from the perspective of immune mechanisms. Here, we present the current understanding of phagocytosis and give a brief discussion of how pathogens control phagocytosis at different stages.

Human-Centric Relation Segmentation: Dataset and Solution.

Vision and language understanding techniques have achieved remarkable progress, but currently it is still difficult to well handle problems involving very fine-grained details. For example, when the robot is told to "bring me the book in the girl's left hand", most existing methods would fail if the girl holds one book respectively in her left and right hand. In this work, we introduce a new task named human-centric relation segmentation (HRS), as a fine-grained case of HOI-det. HRS aims to predict the relations between the human and surrounding entities and identify the relation-correlated human parts, which are represented as pixel-level masks. For the above exemplar case, our HRS task produces results in the form of relation triplets 〈girl [left hand], hold, book 〉 and exacts segmentation masks of the book, with which the robot can easily accomplish the grabbing task. Correspondingly, we collect a new Person In Context (PIC) dataset for this new task, which contains 17,122 high-resolution images and densely annotated entity segmentation and relations, including 141 object categories, 23 relation categories and 25 semantic human parts. We also propose a Simultaneous Matching and Segmentation (SMS) framework as a solution to the HRS task. It contains three parallel branches for entity segmentation, subject object matching and human parsing respectively. Specifically, the entity segmentation branch obtains entity masks by dynamically-generated conditional convolutions; the subject object matching branch detects the existence of any relations, links the corresponding subjects and objects by displacement estimation and classifies the interacted human parts; and the human parsing branch generates the pixelwise human part labels. Outputs of the three branches are fused to produce the final HRS results. Extensive experiments on PIC and V-COCO datasets show that the proposed SMS method outperforms baselines with the 36 FPS inference speed. Notably, SMS outperforms the best performing baseline m-KERN with only 17.6 percent time cost. The dataset and code will be released at http://picdataset.com/challenge/index/.

Diagnostic performance of microRNAs in testicular germ cell tumors: a systematic review and meta-analysis.

The sensitivity (Sen) of classic biomarkers for the diagnosis of testicular germ cell tumors (TGCTs) is currently low. Previous studies have shown the diagnostic potential of microRNAs (miRNAs) for TGCTs; however, the results of these studies are inconsistent. Therefore, we conducted a systematic review and meta-analysis to evaluate their diagnostic value. PubMed, EMBASE, Cochrane Library, and Web of Science databases were systematically searched until September 30, 2020 and 18 trials from 11 studies involving 2,068 participants were included in this meta-analysis. Using a bivariate mixed-effects meta-analysis model, the pooled Sen, specificity (Spe), positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic odds ratio (DOR), and area under the curve (AUC) with 95% confidence interval values of total miRNAs were 0.83 (0.73-0.90), 0.95 (0.89-0.98), 15.79 (7.41-33.66), 0.18 (0.11-0.29), 87.13 (41.99-180.82), and 0.95 (0.93-0.97), respectively; however, the observed values of single miR-371a-3p were 0.84 (0.76-0.90), 0.95 (0.91-0.98), 18.41 (9.69-34.97), 0.17 (0.11-0.26), 111.56 (47.72-260.80), and 0.97 (0.95-0.98), respectively. Subgroup analysis revealed that miRNAs that included miR-371a-3p showed higher predictive performance than those that did not (P < 0.05). This research identified that miR-371a-3p has a high diagnostic value for TGCTs, except teratoma.

Value of Xpert MTB/RIF Using Bronchoalveolar Lavage Fluid for the Diagnosis of Pulmonary Tuberculosis: a Systematic Review and Meta-analysis.

The performance of Xpert MTB/RIF using bronchoalveolar lavage fluid (BAL) for the diagnosis of pulmonary tuberculosis (PTB) remains unclear. Therefore, a systematic review/meta-analysis was conducted. Studies published before 31 December 2019 were retrieved from the PubMed, Embase, and Web of Science databases using the keywords "pulmonary tuberculosis," "Xpert MTB/RIF," and "BAL." Two independent evaluators extracted the data and assessed the bias risk of the included studies. A random-effects model was used to calculate the overall sensitivity, specificity, positive and negative likelihood ratios (PLR and NLR, respectively), diagnostic odds ratio (DOR), and the area under the curve (AUC), as well as the respective 95% confidence intervals (CIs). Nineteen trials involving 3,019 participants met the inclusion criteria. Compared to the culture method, the pooled sensitivity, specificity, PLR, NLR, DOR, and the AUC with 95% CIs of Xpert MTB/RIF were 0.87 (0.84 to 0.90), 0.92 (0.91 to 0.93), 10.21 (5.78 to 18.02), 0.16 (0.12 to 0.22), 78.95 (38.59 to 161.53), and 0.9467 (0.9462 to 0.9472), respectively. Relative to the composite reference standard, the observed values were 0.69 (0.65 to 0.72), 0.98 (0.98 to 0.99), 37.50 (18.59 to 75.62), 0.30 (0.21 to 0.43), 171.98 (80.82 to 365.96), and 0.9691 (0.9683 to 0.9699), respectively. All subgroups, except children, showed high sensitivity and specificity. In conclusion, the use of Xpert MTB/RIF in the context of BAL samples has a high diagnostic performance for PTB (except for children) and may serve as an alternative rapid diagnostic tool.

Systematic review of the registered clinical trials for coronavirus disease 2019 (COVID-19).

BACKGROUND: Since the outbreak of coronavirus disease 2019 (COVID-19), many researchers in China have performed related clinical research. However, systematic reviews of the registered clinical trials are still lacking. Therefore, we conducted a systematic review of clinical trials for COVID-19 to summarize their characteristics. METHODS: This study is based on the PRISMA recommendations in the Cochrane handbook. The Chinese Clinical Registration Center and the ClinicalTrials.gov databases were searched to identify registered clinical trials related to COVID-19. The retrieval inception date was February 9, 2020. Two researchers independently selected the literature based on the inclusion and exclusion criteria, extracted data, and evaluated the risk of bias. RESULTS: A total of 75 registered clinical trials (63 interventional studies and 12 observational studies) for COVID-19 were identified. The majority of clinical trials were sponsored by Chinese hospitals. Only 11 trials have begun to recruit patients, and none of the registered clinical trials have been completed; 34 trials were early clinical exploratory trials or in the pre-experiment stage, 13 trials were phase III, and four trials were phase IV. The intervention methods included traditional Chinese medicine in 26 trials, Western medicine in 30 trials, and integrated traditional Chinese medicine and Western medicine in 19 trials. The subjects were primarily non-critical adult patients (≥ 18 years old). The median sample size of the trials was 100 (IQR: 60-200), and the median length of the trial periods was 179 d (IQR: 94-366 d). The main outcomes were clinical observation and examinations. Overall, the methodological quality of both the interventional trials and observational studies was low. CONCLUSIONS: Intensive clinical trials on the treatment of COVID-19 using traditional Chinese medicine and Western medicine are ongoing or will be performed in China. However, based on the uncertain methodological quality, small sample size, and long trial duration, we will not be able to obtain reliable, high-quality clinical evidence regarding the treatment of COVID-19 in the near future. Improving the quality of study design, prioritizing promising drugs, and using different designs and statistical methods are worth advocating and recommending for clinical trials of COVID-19 in the future.

Use of thiazide diuretics for the prevention of recurrent kidney calculi: a systematic review and meta-analysis.

BACKGROUND: Thiazide diuretics reduce the risk of recurrent kidney calculi in patients with kidney calculi or hypercalciuria. However, whether thiazide diuretics can definitely prevent recurrent kidney calculi remains unclear. We aimed to evaluate the effect and safety of thiazide diuretics on recurrent kidney calculi. METHODS: The PubMed, Cochrane Library, and EMBASE databases were systematically searched using the keywords thiazide diuretics and kidney calculi to identify randomized controlled trials (RCTs). The primary outcome was the incidence of recurrent kidney calculi, and the secondary outcome was the 24-h urinary calcium level. The pooled risk ratio (RR), risk difference (RD), standardized mean difference (SMD), and 95% confidence interval (CI) were calculated. The evidence quality was graded using the GRADE criteria, and recommendations for recurrent kidney calculus prevention using thiazide diuretics were reassessed. RESULTS: Eight RCTs involving 571 patients were included. The pooled RR for the incidence of kidney calculi in the thiazide diuretic groups was 0.44 (95% CI 0.33-0.58, P < 0.0001) compared to that in the placebo and untreated groups; the pooled RD was - 0.23 (95% CI - 0.30 to - 0.16, P < 0.0001). The pooled SMD for the 24-h urinary calcium level was - 18.59 (95% CI - 25.11 to - 12.08, P < 0.0001). The thiazide diuretic groups had a high incidence of adverse reactions and low tolerance. The evidence quality for decrease in kidney calculus incidence using thiazide diuretics was low, while that for the 24-h urinary calcium level decrease among those with recurrent kidney calculi was moderate, and that for the decrease in kidney calculus incidence using short-acting and long-acting thiazide diuretics was low. The overall strength of recommendation for prevention of recurrent renal calculi using thiazide diuretics was not recommended. The subgroup and sensitivity analysis findings were robust. CONCLUSIONS: Long-term use of thiazide diuretics reduces the incidence of recurrent renal calculi and 24-h urinary calcium level. However, the benefits are insufficient, and the evidence quality is low. Considering the adverse effects, poor patient compliance, and economic burden of long-term medication, their use in preventing recurrent kidney calculi is not recommended.

Stimulation of cyclic electron flow around PSI as a response to the combined stress of high light and high temperature in grape leaves.

Changes in cyclic electron flow (CEF) around PSI activity after exposing grape (Vitis vinifera L.) seedling leaves to the combined stress of high temperature (HT) and high light (HL) were investigated. The PSII potential quantum efficiency (Fv/Fm) decreased significantly under exposure to HT, and this decrease was greater when HT was combined with HL, whereas the PSI activity maintained stable. HT enhanced CEF mediated by NAD(P)H dehydrogenase remarkably. Compared with the control leaves, the half-time of P700+ re-reduction decreased during the HT treatment; this decrease was even more pronounced under the combined stress, implying significantly enhanced CEF as a result of the treatment. However, the heat-induced increase in nonphotochemical quenching (NPQ) was greater under HL, accompanied by a greater enhancement in high-energy state quenching. These results suggest that the combined stress of HT and HL resulted in severe PSII photoinhibition, whereas CEF showed plasticity in its response to environmental stress and played an important role in PSII and PSI photoprotection through accelerating generation of the thylakoid proton gradient and the induction of NPQ.

Impact of Hypoxia on Drug Resistance and Growth Characteristics of Mycobacterium tuberculosis Clinical Isolates.

Mycobacterium tuberculosis (MTB) is a specific aerobic bacterium, but can survive under hypoxic conditions, such as those in lung cheese necrosis, granulomas, or macrophages. It is not clear whether the drug sensitivity and growth characteristics of MTB under hypoxic conditions are different from those under aerobic conditions. In this study, we examined the drug resistance and growth characteristics of MTB clinical isolates by a large sample of in vitro drug susceptibility tests, using an automatic growth instrument. Under hypoxic conditions, variance in drug resistance was observed in nearly one-third of the MTB strains and was defined as MTB strains with changed drug sensitivity (MTB-CDS). Among these strains, resistance in a considerable proportion of clinical strains was significantly increased, and some strains emerged as multi-drug resistant. Growth test results revealed a high growth rate and large survival number in macrophages under hypoxia in MTB-CDS. According to the results of fluorescence quantitative PCR, the expression of some genes, including RegX3 (involving RIF resistance), Rv0194 (efflux pump gene), four genes related to transcription regulation (KstR, DosR, Rv0081 and WhiB3) and gene related to translation regulation (DATIN), were upregulated significantly under hypoxic conditions compared to that under aerobic conditions (p < 0.05). Thus, we concluded that some MTB clinical isolates can survive under hypoxic conditions and their resistance could change. As for poor clinical outcomes in patients, based on routine drug susceptibility testing, drug susceptibility tests for tuberculosis under hypoxic conditions should also be recommended. However, the detailed mechanisms of the effect of hypoxia on drug sensitivity and growth characteristics of MTB clinical isolates still requires further study.

High-resolution imaging of labile phosphorus and its relationship with iron redox state in lake sediments.

A thorough understanding of the labile status and dynamics of phosphorus (P) and iron (Fe) across the sediment-water interface (SWI) is essential for managing internal P release in eutrophic lakes. Fe-coupled inactivation of P in sediments is an important factor which affects internal P release in freshwater lakes. In this study, two in-situ high-resolution diffusive gradients in thin films (DGT) techniques, Zr-Oxide DGT and ZrO-Chelex DGT, were used to investigate the release characteristics of P from sediments in a large freshwater lake (Dongting Lake, China; area of 2691 km2) experiencing a regional summer algal bloom. Two-dimensional distributions of labile P in sediments were imaged with the Zr-Oxide DGT without destruction of the original structure of the sediment layer at four sites of the lake. The concentration of DGT-labile P in the sediments, ranging from 0.007 to 0.206 mg L-1, was highly heterogeneous across the profiles. The values of apparent diffusion flux (Fd) and release flux (Fr) of P varied between -0.027-0.197 mg m-2 d-1 and 0.037-0.332 mg m-2 d-1, respectively. Labile P showed a high and positive correlation (p < 0.01) with labile Fe(II) in the profiles, providing high-resolution evidence for the key role of Fe-redox cycling in labile P variation in sediments.

A Systematic Review and Meta-Analysis of the Clinical Appropriateness of Blood Transfusion in China.

The issue of the clinical appropriateness of blood transfusion has become a focus of transfusion medicine worldwide. In China, irrational uses of blood have often been reported in recent years. However, to date there lacks a systematic review of the rational uses of blood. This study aimed to determine the clinical appropriateness of blood transfusion in China. We searched PubMed, Web of Science, the Cochrane Library, China National Knowledge Infrastructure (CNKI), China Science and Technology Journal Database, WanFang Database, and Chinese BioMedical Literature Database, and the retrieval cut-off date was June 31, 2015. SPSS 17.0 and MetaAnalyst 3.13 were employed as the statistics tools in this review. A pooled rate of clinical inappropriateness of transfusion was analyzed by DerSimonian-Laird method. In this study, a total of 39 observational studies were included, which related to 75,132 cases of blood transfusion. According to the meta-analysis results, the overall incidence of clinical inappropriateness of transfusion in China was estimated to be 37.3% (95% confidence interval [CI] [32.1, 42.8]). The subgroup analyses revealed that the pooled rates of clinical inappropriateness of transfusion of plasma, red blood cells (RBCs), cryoprecipitate, and platelets were 56.3% (95% CI [45.8, 66.2]), 30.9% (95% CI [27.1, 35.0]), 25.2% (95% CI [13.2, 42.7]), and 14.1% (95% CI [8.8, 21.9]), respectively. However, the pooled incidence of inappropriateness of transfusion in operative departments was 47.5% (95% CI [36.8, 58.3]), which was significantly higher than that in nonoperative departments, 25.8% (95% CI [18.7, 34.4], P < 0.05). The overall rates of inappropriate use were 36.7% (95% CI [30.2, 43.6]) in major cities and 37.5% (95% CI [31.2, 44.3]) in other cities, respectively; there was no statistically significant difference (P > 0.05). In conclusion, China has suffered from a disadvantage in the clinical appropriateness of blood transfusion, especially in plasma and RBC use. In future, comprehensive measures should be implemented in order to improve the clinical appropriateness of blood transfusion.

Meta-analysis of Huangqi injection for the adjunctive therapy of aplastic anemia.

Aplastic anemia therapy remains difficult, due to lack of effective treatment regimens. In recent years, Huangqi injection for the adjunctive therapy of aplastic anemia has been reported in many clinical trials. Considering that Huangqi injection may be a novel approach to aplastic anemia treatment, we conducted a meta-analysis of clinical controlled trials to assess the clinical value of Huangqi injection in the treatment of aplastic anemia. We searched the Chinese Biomedical Literature Database (CBM), China National Knowledge Infrastructure (CNKI), Chinese Scientific Journals Full-text Database (VIP), Wanfang Database, PubMed and EMBASE database to collect the data about the trials of Huangqi injection combined with androgens for treating aplastic anemia. A total of ten studies involving 720 patients with aplastic anemia were included in this study. The meta-analysis showed significant increases in the pool effectiveness rate, white blood cells (WBC), haematoglobin (Hb), platelets (PLT), and reticulocytes (Ret) between the experimental group versus the control group. No severe side effects were found in this study. However, the lower Jadad scores and asymmetric funnel plot degrades the validity of the meta-analysis as the clinical evidence. Therefore, Huangqi injection may significantly enhance the efficacy of androgens for aplastic anemia, suggesting that the novel approach of Chinese traditional medicine combined with Western medicine is promising. The exact outcome required confirmation with rigorously well-designed multi-center trials.

Research of the heteroresistance of Pseudomonas aeruginosa to imipenem.

Pseudomonas aeruginosa (PA) plays plays an important role in nosocomial infection. To explore the heteroresistance of PA to imipenem (IMP), we detected the sensitivity of 140 strains of PA to IMP using the KB method and VITEK method. Combined with resistance mutation analysis, the heteroresistance of PA to IMP was determined. Whilst, the double disk synergy test and SYBGreen RT-PCR for efflux pump were performed in IMP-heteroresistant PA strains. In this study, we confirmed 20 IMP-heteroresistant strains. The double disk synergy tests suggested that none of 20 heteroresistant PA strains produced metalloenzyme. The SYBGreen quantitative RT-PCR revealed that the MexAB expression level of efflux pump in IMP-heteroresistant PA was significantly higher than that in the IMP-sensitive strains (P<0.05), while there was no significantly different between the MexCD expression between resistant strains and sensitive strains (P<0.05). We believe that the clinicians should pay more attention to the PA heteroresistance to IMP, and the heteroesistance of PA to IMP is related to high expression in the MexAB of PA efflux pump.

Better plans and more powerful evidence are needed in the research and treatment of chronic hepatitis B in China.

China has the first largest global pool of chronic hepatitis B (CHB) infection. With the implementation of comprehensive intervention of neonatal vaccination based on national strategy, some achievements have been made in control of hepatitis B in China. However, some problems remain unresolved in the research field as well as management of CHB. Currently, empirical therapy, poor study design, inefficient management, low compliance rate of patients, lack of public awareness and economic factor hampers the research and treatment of CHB in China. We propose that China should take more effective measures to improve the study quality, social management, the relief of economic burden for the needed patients, public awareness, evidence-based policies and medicine, supervision for clinical practice, diagnostics ability and laboratory quality, the compliance of patients, and the quality of Traditional Chinese Medicine (TCM) in the management of CHB infection in the future.