Search

In this work, we present a comprehensive study on the use of deep neural networks (DNNs) for automatic language identification (LID). Motivated by the recent success of using DNNs in acoustic modeling for speech recognition, we adapt DNNs to the problem of identifying the language in a given utterance from its short-term acoustic features. We propose two different DNN-based approaches. In the first one, the DNN acts as an end-to-end LID classifier, receiving as input the speech features and providing as output the estimated probabilities of the target languages. In the second approach, the DNN is used to extract bottleneck features that are then used as inputs for a state-of-the-art i-vector system. Experiments are conducted in two different scenarios: the complete NIST Language Recognition Evaluation dataset 2009 (LRE'09) and a subset of the Voice of America (VOA) data from LRE'09, in which all languages have the same amount of training data. Results for both datasets demonstrate that the DNN-based systems significantly outperform a state-of-art i-vector system when dealing with short-duration utterances. Furthermore, the combination of the DNN-based and the classical i-vector system leads to additional performance improvements (up to 45% of relative improvement in both EER and C a v g on 3s and 10s conditions, respectively). [ABSTRACT FROM AUTHOR]

This article proposes an adaptable hybrid model for recommending effective investments in different scenarios. Currently, a wide variety of methodologies are used for company valuation, especially those that take into account financial statements. However, for private held companies, there is no method that would be capable of predicting, with full certainty, the future success of an investment. The Optimal Capital Investment Case-Base Reasoning (OCI-CBR) consists of a case-based reasoning system that uses a classification algorithm to prune the case base according to a projected increase in certain company attributes. Once the cases have been pruned and the case is fed with the most profitable investment opportunities, the case-based reasoning system recommends optimal investments to potential investors. The complete model is conceived as an intelligent hybrid model that optimizes the case base by employing different algorithms for data retrieval and reuse. The system makes recommendations based on the investor's preferences and the investment decisions of other investors with similar profiles or interests. • Hybrid decision support system for capital investment. • Machine learning pruning techniques for fault detection and diagnosis of companies. • A CBR methodology is used to make recommendations for optimal investment. [ABSTRACT FROM AUTHOR]

This work addresses the use of deep neural networks (DNNs) in automatic language identification (LID) focused on short test utterances. Motivated by their recent success in acoustic modelling for speech recognition, we adapt DNNs to the problem of identifying the language in a given utterance from the short-term acoustic features. We show how DNNs are particularly suitable to perform LID in real-time applications, due to their capacity to emit a language identification posterior at each new frame of the test utterance. We then analyse different aspects of the system, such as the amount of required training data, the number of hidden layers, the relevance of contextual information and the effect of the test utterance duration. Finally, we propose several methods to combine frame-by-frame posteriors. Experiments are conducted on two different datasets: the public NIST Language Recognition Evaluation 2009 (3 s task) and a much larger corpus (of 5 million utterances) known as Google 5M LID, obtained from different Google Services. Reported results show relative improvements of DNNs versus the i -vector system of 40% in LRE09 3 second task and 76% in Google 5M LID. [ABSTRACT FROM AUTHOR]

Abstract: Heart rate variability (HRV) is a complex signal that results from the contribution of different sources of oscillation related to the autonomic nervous system activity. Although linear analysis of HRV has been applied to sleep studies, the nonlinear dynamics of HRV underlying frequency components during sleep is less known. We conducted a study to evaluate nonlinear HRV within independent frequency components in wake status, slow-wave sleep (SWS, stages III or IV of non-rapid eye movement sleep), and rapid-eye-movement sleep (REM). The sample included 10 healthy adults. Polysomnography was performed to detect sleep stages. HRV was studied globally during each phase and then very low frequency (VLF), low frequency (LF) and high frequency (HF) components were separated by means of the wavelet transform algorithm. HRV nonlinear dynamics was estimated with sample entropy (SampEn). A higher SampEn was found when analyzing global variability (Wake: 1.53±0.28, SWS: 1.76±0.32, REM: 1.45±0.19, p =0.005) and VLF variability (Wake: 0.13±0.03, SWS: 0.19±0.03, REM: 0.14±0.03, p<0.001) at SWS. REM was similar to wake status regarding nonlinear HRV. We propose nonlinear HRV is a useful index of the autonomic activity that characterizes the different sleep–wake cycle stages. [Copyright &y& Elsevier]

Background: This study evaluates the diagnostic performance of high-sensitivity troponin using point-of-care testing (POCT) devices compared with main laboratory measurements for ruling out non-ST-elevation myocardial infarction (NSTEMI) in emergency department (ED) patients presenting with non-traumatic chest pain. Methods: This multicenter, observational, prospective, non-interventional study was conducted in two Spanish hospitals from 1 June to 31 December 2023 and included adult patients presenting with non-traumatic chest pain admitted to the ED. High-sensitivity troponin levels were measured using both the Siemens Atellica® VTLi POCT device and main laboratory testing, with data collected on analytical results and measurement times. Results: Of the 201 patients who met the inclusion criteria, a significant correlation was observed between the POCT and laboratory assays. The area under the curve (AUC) of the ROC curve was consistently greater than 0.9, indicating a high diagnostic accuracy for ruling out NSTEMI. In addition, measurement times were significantly reduced using POCT compared to the core laboratory. Conclusion: These results suggest that high-sensitivity troponin POCT devices offer comparable diagnostic performance to traditional laboratory methods for the diagnosis of NSTEMI in the emergency department, potentially speeding up clinical decisions and optimizing resource utilization. [ABSTRACT FROM AUTHOR]

Background: Myopia prevalence has been increasing in the last decades, and its pathological consequences, including myopic maculopathy and high myopia-associated optic neuropathy, are now one of the most common causes of visual impairment. It is estimated that by 2050, more than 50% of Europeans and Americans will be myopes, which is alarming due to the high morbidity of myopes over − 6.00D. Once myopia has appeared, there are different options with scientific evidence to try to slow the axial length growth. Ophthalmic lenses are the less invasive treatment to control myopia, and there is evidence about the efficacy of different designs, mainly in the Asiatic population. However, new designs have been launched, and it is not known if efficacy is the same between Asiatic and European subjects. Thus, we have set up a randomized, controlled, double-blind, and multicenter trial to investigate the efficacy of a new design of ophthalmic lenses for myopia control in European children. Methods: A 2-year prospective, multicenter, randomized controlled, and double-blind clinical trial is used to investigate the efficacy of a new design of ophthalmic lenses to slow the progression of myopia. Three hundred children aged from 6 to 13 years old will be recruited and randomly assigned to a study or control group. The study group will be composed of 150 children wearing MyoCare while the control group will be composed of 150 children wearing Clearview. The inclusion criteria will be myopia with a spherical equivalent between − 0.75D and − 5.00D, astigmatism < 1.50D, and anisometropia < 1.00D and having a historical evolution of at least − 0.50 The primary outcome is to compare the mean annual progression of the spherical equivalent between both groups. The secondary outcomes are axial length, choroidal thickness, phorias, and accommodative status of both groups. Discussion: This study will be the first randomized and controlled clinical trial in European children with spectacle lenses based on simultaneous competing defocus. The results will shed light on the clinical evidence of spectacle lenses relying on this new design for the management of myopia with results of efficacy in the non-Asiatic population. Trial registration: EU Clinical Trials Register (EudraCT) 2022–001696. Registered on 27 April 2022. ClinicalTrials.gov NCT05919654. Registered on 26 June 2023. [ABSTRACT FROM AUTHOR]

Long Short Term Memory (LSTM) Recurrent Neural Networks (RNNs) have recently outperformed other state-of-the-art approaches, such as i-vector and Deep Neural Networks (DNNs), in automatic Language Identification (LID), particularly when dealing with very short utterances (∼3s). In this contribution we present an open-source, end-to-end, LSTM RNN system running on limited computational resources (a single GPU) that outperforms a reference i-vector system on a subset of the NIST Language Recognition Evaluation (8 target languages, 3s task) by up to a 26%. This result is in line with previously published research using proprietary LSTM implementations and huge computational resources, which made these former results hardly reproducible. Further, we extend those previous experiments modeling unseen languages (out of set, OOS, modeling), which is crucial in real applications. Results show that a LSTM RNN with OOS modeling is able to detect these languages and generalizes robustly to unseen OOS languages. Finally, we also analyze the effect of even more limited test data (from 2.25s to 0.1s) proving that with as little as 0.5s an accuracy of over 50% can be achieved. [ABSTRACT FROM AUTHOR]

Abstract: The recent statements of both the European Union and the US Presidency pushed in the direction of using renewable forms of energy, in order to act against climate changes induced by the growing concentration of carbon dioxide in the atmosphere. In this paper, a survey regarding methods and tools presently available to determine potential and exploitable energy in the most important renewable sectors (i.e., solar, wind, wave, biomass and geothermal energy) is presented. Moreover, challenges for each renewable resource are highlighted as well as the available tools that can help in evaluating the use of a mix of different sources. [ABSTRACT FROM AUTHOR]

Abstract: Porcine circovirus type 2 (PCV2) is a single-stranded circular DNA virus infecting domestic pigs worldwide. Interaction of this virus with the immune system apparently modulates the immune response of the host. In the present study, the implication of different components of PCV2 in the modulation of the immune response of the host were investigated by using PCV2 viral-like particles (VLPs) and 16 novel oligodeoxyribonucleotides containing CpG motifs (CpG-ODNs) based on the PCV2 genomic sequence. The role of these viral components was studied by evaluating the cytokine profiles (IFN-α, IFN-γ, IL-10, IL-2 and IL-12) on porcine peripheral mononuclear cell (PBMC) and bone marrow-derived dendritic cell (BMDC) cultures. Also, the effect of PCV2 and its elements were examined in recall antigen (pseudorabies virus, PRV) responses. While PCV2 was a potent inducer of IL-10 by PBMCs, such effect was not observed using CpG-ODNs or VLPs. However, IFN-γ and IL-2 production by recall antigen was repressed in presence of PCV2 and most of the studied CpG-ODNs. VLPs did not have such repressive effect. In BMDC cultures, PCV2 and most of CpG-ODNs were able to inhibit IFN-α secretion induced by PRV. Interestingly, CpG-ODNs with inhibitory effect were located within the PCV2 Rep gene. Additionally, PCV2 virus was a very strong IL-12 inducer in BMDC cultures. Whereas, IFN-α modulation on BMDC after PCV2 VLP treatment was neglectable, PCV2 VLPs were potent IL-12 inducers. Our data shows that PCV2 viral elements can distinctly regulate cytokine production depending on the cell population studied. Thus, the final immune response upon PCV2 infection seems to depend on the fine balance between the regulatory elements present in viral DNA and structural protein within the host immune system. [Copyright &y& Elsevier]

Background: Mexico is on the top five countries with the highest number of TB cases in America continent, nevertheless, information about genotypes circulating is practically unknown. Considering the above this study aims to characterize the genetic diversity of TB in the city of Veracruz, México.Methods: A cross-sectional study was conducted among positive smear samples from patients living in Veracruz City, samples were cultured, and first-line drug profiles determined. Genotyping was made by spoligotyping and MIRU-VNTR 24 loci. Associations of lineages, clusters, and variables were also analyzed.Results: Among the 202 isolates analyzed resistance to at least one drug was observed in 60 (30%) isolates and 41(20%) were multidrug-resistant. Three major lineages were identified: L4/Euro-American (88%), L1/Indo-Oceanic (9%), and L2/East Asian (3%). The Euro-American lineage included more than six sublineages, the most abundant were: H (32%), T (23%), LAM (18%), and X (12%). 140 isolates (70%) were placed in 42 SITs patterns.Conclusions: These results provide the first baseline data on the genetic structure of TB in the city of Veracruz. Sublineages H, X and LAM were predominant; however, it was founded an important diversity of genotypes that could contribute to the dispersion of TB and explain the high prevalence. This information might be useful for the development of further interventions to reduce impact of TB. [ABSTRACT FROM AUTHOR]

Tuberculosis is still a global public health problem, with an estimated 10 million new cases and 1.6 million deaths in 2017. Of all humans infected with M. tuberculosis, only 10-15% will develop active tuberculosis disease during their lifetime, and data suggest that along with environmental factors, genetic factors influence susceptibility to develop active disease. Toll-like receptors (TLRs) are pattern recognition receptors that play a central role in the initiation and shaping of adaptive immune responses, and several TLRs have been shown to recognize mycobacterial components. In this work, we performed a case-control study to determine if common single nucleotide polymorphisms (SNPs) in genes encoding TLRs 1, 2, 4, 6, and 10 are associated with susceptibility to develop active tuberculosis in population from the state of Veracruz, Mexico. The study included 279 cases and 569 controls. The results show that the frequency of two SNPs in TLR4 was significantly higher in controls than in tuberculosis patients. The minor allele (G) of rs4986790 in TLR4 (D299G) decreased the risk of active tuberculosis in the allelic (A vs. G, OR = 0.31, 95%CI = 0.09-0.81, p = 0.01) and in the dominant genetic model (AA vs. GG+AG, OR = 0.26, 95%CI = 0.09-0.77, p = 0.02). Similarly, the minor allele (T) of rs4986791 in TLR4 (T399I) decreased the risk of active disease in the allelic model (C vs. T, OR = 0.29, 95%CI = 0.10-0.90, p = 0.03). We did not find an association of SNPs in TLR1 (N248S), TLR2 (R753Q), TLR6 (S249P), and TLR10 (A153S and V298I) with tuberculosis disease. These results suggest that in this population, genetic variants of TLR4 affect the susceptibility for suffering active tuberculosis disease. [ABSTRACT FROM AUTHOR]

Hydrogen production from renewable energies is a key part in the energy transition to realize a sustainable energy economy for both developed and developing nations. For Algeria, successful energy transition toward a hydrogen economy will require the establishment of its potential. This study was conducted to estimate the potential for producing hydrogen from renewable resources in Algeria. The renewable energies considered are: solar photovoltaic and wind. To accomplish this objective, first, we analyzed renewable resource data both statistically and graphically using Geographical Information System (GIS), a computer-based information system utilized to create and visualize the spatial distribution of the geographic information. Then, the study will evaluate the availability of renewable electricity production potential from these key renewable resources. The potential for the hydrogen production, via the electrolysis process with wind and solar photovoltaic electricity, is described with maps showing it per unit area in each region. Finally, the results of the estimated hydrogen potential from both resources for each region are compared and significant conclusions are drawn. [ABSTRACT FROM AUTHOR]

The main cellular target for African swine fever virus (ASFV) is the porcine macrophage. However, existing data about the early phases of infection were previously characterized in non-leukocyte cells such as Vero cells. Here, we report that ASFV enters the natural host cell using dynamin-dependent and clathrin-mediated endocytosis. This pathway is strongly pH-dependent during the first steps of infection in porcine macrophages. We investigated the effect of drugs inhibiting several endocytic pathways in macrophages and compared ASFV with vaccinia virus (VV), which apparently involves different entry pathways. The presence of cholesterol in cellular membranes was found to be essential for a productive ASFV infection while actin-dependent endocytosis and the participation of phosphoinositide-3-kinase (PI3K) activity were other cellular factors required in the process of viral entry. These findings improved our understanding of the ASFV interactions with macrophages that allow for successful viral replication. [ABSTRACT FROM AUTHOR]

Abstract: Swine influenza virus (SwIV) is considered a zoonosis and the fact that swine may act as an intermediate reservoir for avian influenza virus, potentially infectious for humans, highlights its relevance and the need to understand the interaction of different influenza viruses with the porcine immune system. Thus, in vitro porcine bone marrow-derived dendritic cell (poBMDCs) were infected with a circulating SwIV A/Swine/Spain/SF32071/2007(H3N2), 2009 human pandemic influenza virus A/Catalonia/63/2009(H1N1), low pathogenic avian influenza virus (LPAIV) A/Anas plathyrhynchos/Spain/1877/2009(aH7N2) or high pathogenic avian influenza virus (HPAIV) A/Chicken/Italy/5093/1999(aH7N1). Swine influenza virus H3N2 infection induced an increase of SLA-I and CD80/86 at 16 and 24h post infection (hpi), whereas the other viruses did not. All viruses induced gene expression of NF-κB, TGF-β, IFN-β and IL-10 at the mRNA level in swine poBMDCs to different extents and in a time-dependent manner. All viruses induced the secretion of IL-12 mostly at 24hpi whereas IL-18 was detected at all tested times. Only swH3N2 induced IFN-α in a time-dependent manner. Swine H3N2, aH7N2 and aH7N1 induced secretion of TNF-α also in a time-dependent manner. Inhibition of NF-κB resulted in a decrease of IFN-α and IL-12 secretion by swH3N2-infected poBMDC at 24hpi, suggesting a role of this transcription factor in the synthesis of these cytokines. Altogether, these data might help in understanding the relationship between influenza viruses and porcine dendritic cells in the innate immune response in swine controlled through soluble mediators and transcription factors. [Copyright &y& Elsevier]

Background Patients in fulminant hepatic failure currently do not have a temporary means of support while awaiting liver transplantation. A potential therapeutic approach for such patients is the use of extracorporeal perfusion with porcine livers as a form of 'liver dialysis'. During a 72-h extracorporeal perfusion of porcine livers with human blood, porcine Kupffer cells bind to and phagocytose human red blood cells (h RBC) causing the hematocrit to decrease to 2.5% of the original value. Our laboratory has identified porcine sialoadhesin expressed on Kupffer cells as the lectin responsible for binding N-acetylneuraminic acid on the surface of the h RBC. We evaluated whether blocking porcine sialoadhesin prevents the recognition and subsequent destruction of h RBCs seen during extracorporeal porcine liver xenoperfusion. Methods Ex vivo studies were performed using wild type pig livers perfused with isolated h RBCs for 72-h in the presence of an anti-porcine sialoadhesin antibody or isotype control. Results The addition of an anti-porcine sialoadhesin antibody to an extracorporeal porcine liver xenoperfusion model reduces the loss of h RBC over a 72-h period. Sustained liver function was demonstrated throughout the perfusion. Conclusions This study illustrates the role of sialoadhesin in mediating the destruction of h RBCs in an extracorporeal porcine liver xenoperfusion model. [ABSTRACT FROM AUTHOR]

The present study describes the distinct bone marrow (BM) and peripheral blood (PB) monocyte subpopulations detected by seven-colour flow cytometry. Mononuclear phagocytes were identified as viable CD172a+ SWC8- CD203a- mononuclear leukocytes. After that, monocyte subpopulations were differentiated by using CD14, CD163 and SLA-DR markers. Four distinct monocyte subpopulations were found in the BM and PB. Based on the discovered populations two possible maturation pathways have been proposed. The first pathway was characterised by release of CD14hi CD163- SLA-DR- BM monocytes into the PB where they matured into CD14low CD163+ SLA-DR+ monocytes. In the alternative pathway the monocytes finalised their phenotypical maturation in the BM and then they were released into the PB as CD14low CD163+ SLA-DR+ cells. In Salmonella-infected piglets, the population of CD14low CD163+ SLA-DR+ monocytes was elevated in the BM and mesenteric lymph nodes (MLN), suggesting the role of this population in pathogenesis of Salmonella infection in pigs. [ABSTRACT FROM AUTHOR]

Introduction and Objectives: We report the results of a randomized controlled trial comparing three different lithotriptors using semirigid ureteroscopy (URS) for distal ureteral stones. Methods: Between September 2009 and November 2010 69 patients undergoing ureteroscopy were randomized to three groups: LithoClast classic (Group 1), Holmium Laser (Group 2), and StoneBreaker™ (Group 3). A 7.5F semirigid ureteroscope was used in all procedures. The primary outcome was differences in fragmentation time. Secondary outcomes were stone-free rates, intraoperative complications, stone-up migration, hospital stay, analgesic requirement, and need for auxiliary procedures. Patients were followed up at 15 days, 30 days, and 3 months. The stone-free status was defined with noncontrast computed tomography performed at first control. Univariate and multivariate analysis were performed to determine clinical and surgical factors that have direct impact on the success of ureteroscopy. Chi-square test and Analysis of Covariance (ANCOVA) tests were used for statistical comparisons. Results: There were no differences between sociodemographic variables. Average stone size was 7.17±2.04 mm in Group 1; 7.89±2.73 mm in Group 2; and 7.79±2.97 mm in Group 3 ( p=0.79). Fragmentation time were similar between lithotriptors; 27.12±4.07 minutes in Lithoclast group; 21.78±2.81 minutes in Laser group, and 27.14±4.71 minutes in StoneBreaker group ( p=0.74). Stone-free rates were 96%±11.18% (group 1), 96.9%±8% (group 2), and 96.9%±8.4% (group 3) ( p=0.1). No difference was observed in stone-up migration, postoperative Double-J stent placement, or auxiliary procedures. Stone size and the placement of a second working wire were associated with shorter fragmentation time ( p<0.01). Conclusions: The three lithotripsy devices evaluated behaved similarly in terms of the ability to fragment stones, and were equally effective for distal ureteral stones. Adequate fragmentation and fragment removal are mainly dependant on stone size and surgical technique (use of auxiliary wire). [ABSTRACT FROM AUTHOR]

Abstract: Swine influenza virus (SwIV) causes sub-acute or acute respiratory infections on swine farms and pigs can act as “mixing vessels” for new influenza strains. Knowledge of the immune response of SwIV in its natural host, pigs, is very limited. Dendritic cells (DCs) mediate the induction of immunity to pathogens, but their interaction with SwIV has not been fully characterized. Thus, porcine bone marrow derived DCs (poBMDCs) were exposed to a circulating strain of H3N2 SwIV in vitro. Infection of poBMDCs resulted in structures resembling influenza virus inside poBMDCs in vesicles and also free in cytoplasm. Viral progeny was undetectable in supernatant but limited replication was detected in the first 8h after infection. However, viral particles from infected-poBMDCs were able to induce cytopathic effect in susceptible cells only when cell-to-cell interaction was favoured. The data generated in our studies reveal the particular interaction of H3N2 SwIV with conventional DCs. [Copyright &y& Elsevier]

Abstract: Development of efficient and safer vaccines against foot-and-mouth disease virus (FMDV) is a must. Previous results obtained in our laboratory have demonstrated that DNA vaccines encoding B and T cell epitopes from type C FMDV, efficiently controlled virus replication in mice, while they did not protect against FMDV challenge in pigs, one of the FMDV natural hosts. The main finding of this work is the ability to improve the protection afforded in swine using a new DNA-vaccine prototype (pCMV-APCH1BTT), encoding FMDV B and T-cell epitopes fused to the single-chain variable fragment of the 1F12 mouse monoclonal antibody that recognizes Class-II Swine Leukocyte antigens. Half of the DNA-immunized pigs were fully protected upon viral challenge, while the remaining animals were partially protected, showing a delayed, shorter and milder disease than control pigs. Full protection in a given vaccinated-pig correlated with the induction of specific IFNγ-secreting T-cells, detectable prior to FMDV-challenge, together with a rapid development of neutralizing antibodies after viral challenge, pointing towards the relevance that both arms of the immune response can play in protection. Our results open new avenues for developing future FMDV subunit vaccines. [Copyright &y& Elsevier]

Abstract: The objective of this work was to explore whether a plasmid expressing CCL20 chemokine could improve the immune response against CSFV in co-administration with a DNA vaccine expressing the E2 protein. The immunization of pigs with the DNA vaccine formulation, that contains swine CCL20 chemokine, resulted in the homogenous induction of detectable levels of CSFV antibodies at 36 days after the first injection. Remarkably, immunized animals with E2 DNA vaccine in co-administration with the plasmid containing swine CCL20 developed high titers of neutralizing antibodies against homologous and heterologous CSFV strains and were totally protected upon a lethal viral challenge (sterilizing protection). Our results confirm the role of CCL20 to increase antibody-mediated responses. At the same time suggest the ability of CCL20 to enhance the T helper cell response associated with the induction of neutralizing antibodies against CSFV in pigs previously reported. Systemic replication of virulent CSFV in vivo during the acute phase of infection induces type I IFN. Lower average values of IFN alpha were detected in the serum of pigs immunized with pE2 and pCCL20 at 3 days after challenge. The levels of IFN-alpha detected in pigs immunized with pE2 and principally in non-vaccinated challenged animals can be related to viral load in serum at 3 and 7 days post infection and the clinical signs observed. Our results emphasized the capacity of swine CCL20 chemokine to enhance cellular, humoral and anti viral response with an adjuvant effect in the immune response elicited by E2-DNA vaccination against CSFV. To our knowledge, this is the first report demonstrating the adjuvant effect of swine CCL20 to effectively enhance the potential of DNA vaccine in the immune induction and protection against virus challenge in swine infection model. [Copyright &y& Elsevier]

Swine monocytes constitute a heterogeneous population of cells which can be divided into four subsets based on the expression of SWC3, CD14, CD163 and swine leucocyte antigen (SLA) DR markers. These subsets appear to represent different maturation stages in a pathway along which these cells up-regulate the expression of SLA DR and CD163 antigens and reduce that of CD14. Differences in the expression of adhesion and costimulatory molecules are also patent, with a progressive increase in the expression of CD11a, wCD11R1, CD29, CD49d, CD61, CD1a and CD80/86, and a concomitant decrease in that of wCD11R2. Besides, these subsets differ in their capacity for tumour necrosis factor-α (TNF-α) production in response to lipopolysaccharide + interferon-γ. The CD163+ CD14− SLA DR+ subset produces higher amounts of TNF-α than the CD163− CD14+ SLA DR− subset, whereas CD163+ CD14+ SLA DR+ and CD163− CD14+ SLA DR+ subsets show intermediate values. CD163+ monocytes also display a higher ability to present soluble antigens to T cells than CD163− monocytes. [ABSTRACT FROM AUTHOR]