Parallel GPU architecture for hyperspectral unmixing based on augmented Lagrangian method

Hyperspectral imaging has become one of the main topics in remote sensing applications, which comprise hundreds of spectral bands at different (almost contiguous) wavelength channels over the same area generating large data volumes comprising several GBs per flight. This high spectral resolution can be used for object detection and for discriminate between different objects based on their spectral characteristics. One of the main problems involved in hyperspectral analysis is the presence of mixed pixels, which arise when the spacial resolution of the sensor is not able to separate spectrally distinct materials. Spectral unmixing is one of the most important task for hyperspectral data exploitation. However, the unmixing algorithms can be computationally very expensive, and even high power consuming, which compromises the use in applications under on-board constraints. In recent years, graphics processing units (GPUs) have evolved into highly parallel and programmable systems. Specifically, several hyperspectral imaging algorithms have shown to be able to benefit from this hardware taking advantage of the extremely high floating-point processing performance, compact size, huge memory bandwidth, and relatively low cost of these units, which make them appealing for onboard data processing. In this paper, we propose a parallel implementation of an augmented Lagragian based method for unsupervised hyperspectral linear unmixing on GPUs using CUDA. The method called simplex identification via split augmented Lagrangian (SISAL) aims to identify the endmembers of a scene, i.e., is able to unmix hyperspectral data sets in which the pure pixel assumption is violated. The efficient implementation of SISAL method presented in this work exploits the GPU architecture at low level, using shared memory and coalesced accesses to memory.

Unsupervised hyperspectral signal subspace identification

Hyperspectral imaging sensors provide image data containing both spectral and spatial information from the Earth surface. The huge data volumes produced by these sensors put stringent requirements on communications, storage, and processing. This paper presents a method, termed hyperspectral signal subspace identification by minimum error (HySime), that infer the signal subspace and determines its dimensionality without any prior knowledge. The identification of this subspace enables a correct dimensionality reduction yielding gains in algorithm performance and complexity and in data storage. HySime method is unsupervised and fully-automatic, i.e., it does not depend on any tuning parameters. The effectiveness of the proposed method is illustrated using simulated data based on U.S.G.S. laboratory spectra and real hyperspectral data collected by the AVIRIS sensor over Cuprite, Nevada.

Identification of factors and groups at risk of infection with Schistosoma mansoni: a strategy for the implementation of control measures?

A fourteen year schistosomiasis control program in Peri-Peri (Capim Branco, MG) reduced prevalence from 43.5 to 4.4%; incidence from 19.0 to 2.9%, the geometric mean of the number of eggs from 281 to 87 and the level of the hepatoesplenic form cases from 5.9 to 0.0%. In 1991, three years after the interruption of the program, the prevalence had risen to 19.6%. The district consists of Barbosa (a rural area) and Peri-Peri itself (an urban area). In 1991, the prevalence in the two areas was 28.4% and 16.0% respectively. A multivariate analysis of risk factors for schistosomiasis indicated the domestic agricultural activity with population attributive risk (PAR) of 29.82%, the distance (< 10 m) from home to water source (PAR = 25.93%) and weekly fishing (PAR = 17.21%) as being responsible for infections in the rural area. The recommended control measures for this area are non-manual irrigation and removal of homes to more than ten meters from irrigation ditches. In the urban area, it was observed that swimming at weekly intervals (PAR = 20.71%), daily domestic agricultural activity (PAR = 4.07%) and the absence of drinking water in the home (PAR=4.29%) were responsible for infections. Thus, in the urban area the recommended control measures are the substitution of manual irrigation with an irrigation method that avoids contact with water, the creation of leisure options of the population and the provision of a domestic water supply. The authors call attention to the need for the efficacy of multivariate analysis of risk factors to be evaluated for schistosomiasis prior to its large scale use as a indicator of the control measures to be implemented.

Parallel hyperspectral unmixing method via split augmented lagrangian on GPU

One of the main problems of hyperspectral data analysis is the presence of mixed pixels due to the low spatial resolution of such images. Linear spectral unmixing aims at inferring pure spectral signatures and their fractions at each pixel of the scene. The huge data volumes acquired by hyperspectral sensors put stringent requirements on processing and unmixing methods. This letter proposes an efficient implementation of the method called simplex identification via split augmented Lagrangian (SISAL) which exploits the graphics processing unit (GPU) architecture at low level using Compute Unified Device Architecture. SISAL aims to identify the endmembers of a scene, i.e., is able to unmix hyperspectral data sets in which the pure pixel assumption is violated. The proposed implementation is performed in a pixel-by-pixel fashion using coalesced accesses to memory and exploiting shared memory to store temporary data. Furthermore, the kernels have been optimized to minimize the threads divergence, therefore achieving high GPU occupancy. The experimental results obtained for the simulated and real hyperspectral data sets reveal speedups up to 49 times, which demonstrates that the GPU implementation can significantly accelerate the method's execution over big data sets while maintaining the methods accuracy.

Comparison of Disposable Pipette Extraction and Dispersive Solid-Phase Extraction in the QuEChERS Method for Analysis of Pesticides in Strawberries

In this study, we sought to assess the applicability of GC–MS/MS for the identification and quantification of 36 pesticides in strawberry from integrated pest management (IPM) and organic farming (OF). Citrate versions of QuEChERS (quick, easy, cheap, effective, rugged and safe) using dispersive solid-phase extraction (d-SPE) and disposable pipette extraction (DPX) for cleanup were compared for pesticide extraction. For cleanup, a combination of MgSO4, primary secondary amine and C18 was used for both the versions. Significant differences were observed in recovery results between the two sample preparation versions (DPX and d-SPE). Overall, 86% of the pesticides achieved recoveries (three spiking levels 10, 50 and 200 µg/kg) in the range of 70–120%, with <13% RSD. The matrix effects were also evaluated in both the versions and in strawberries from different crop types. Although not evidencing significant differences between the two methodologies were observed, however, the DPX cleanup proved to be a faster technique and easy to execute. The results indicate that QuEChERS with d-SPE and DPX and GC–MS/MS analysis achieved reliable quantification and identification of 36 pesticide residues in strawberries from OF and IPM.

EVALUATION OF A RAPID SCREENING ASSAY FOR BACTERIAL IDENTIFICATION (DOT-ELISA) IN FECAL SAMPLES FROM CHILDREN

With the objective of standardizing a Dot Enzyme-Linked Immunosorbent Assay (Dot-ELISA) to detect antigens of fecal bacterial enteropathogens, 250 children, aged under 36 months and of both sexes, were studied; of which 162 had acute gastroenteritis. The efficacy of a rapid screening assay for bacterial enteropathogens (enteropathogenic Escherichia coli "EPEC", enteroinvasive Escherichia coli "EIEC", Salmonella spp. and Shigella spp.) was evaluated. The fecal samples were also submitted to a traditional method of stool culture for comparison. The concordance index between the two techniques, calculated using the Kappa (k) index for the above mentioned bacterial strains was 0.8859, 0.9055, 0.7932 and 0.7829 respectively. These values express an almost perfect degree of concordance for the first two and substantial concordance for the latter two, thus enabling this technique to be applied in the early diagnosis of diarrhea in infants. With a view to increasing the sensitivity and specificity of this immunological test, a study was made of the antigenic preparations obtained from two types of treatment: 1) deproteinization by heating; 2) precipitation and concentration of the lipopolysaccharide antigen (LPS) using an ethanol-acetone solution, which was then heated in the presence of sodium EDTA

hsp65 PCR-restriction enzyme analysis (PRA) for identification of mycobacteria in the clinical laboratory

More than 70 species of mycobacteria have been defined, and some can cause disease in humans, especially in immunocompromised patients. Species identification in most clinical laboratories is based on phenotypic characteristics and biochemical tests and final results are obtained only after two to four weeks. Quick identification methods, by reducing time for diagnosis, could expedite institution of specific treatment, increasing chances of success. PCR restriction-enzyme analysis (PRA) of the hsp65 gene was used as a rapid method for identification of 103 clinical isolates. Band patterns were interpreted by comparison with published tables and patterns available at an Internet site (http://www.hospvd.ch:8005). Concordant results of PRA and biochemical identification were obtained in 76 out of 83 isolates (91.5%). Results from 20 isolates could not be compared due to inconclusive PRA or biochemical identification. The results of this work showed that PRA could improve identification of mycobacteria in a routine setting because it is accurate, fast, and cheaper than conventional phenotypic identification.

Molecular identification of Candida dubliniensis isolated from oral lesions of HIV-positive and HIV-negative patients in São Paulo, Brazil

Candida dubliniensis is a new, recently described species of yeast. This emerging oral pathogen shares many phenotypic and biochemical characteristics with C. albicans, making it hard to differentiate between them, although they are genotypically distinct. In this study, PCR (Polymerase Chain Reaction) was used to investigate the presence of C. dubliniensis in samples in a culture collection, which had been isolated from HIV-positive and HIV-negative patients with oral erythematous candidiasis. From a total of 37 samples previously identified as C. albicans by the classical method, two samples of C. dubliniensis (5.4%) were found through the use of PCR. This study underscores the presence of C. dubliniensis, whose geographical and epidemiological distribution should be more fully investigated.

Primary screening of blood donors by nat testing for HCV-RNA: development of an "in-house" method and results

An "in-house" RT-PCR method was developed that allows the simultaneous detection of the RNA of the Hepatitis C Virus (HCV) and an artificial RNA employed as an external control. Samples were analyzed in pools of 6-12 donations, each donation included in two pools, one horizontal and one vertical, permitting the immediate identification of a reactive donation, obviating the need for pool dismembering. The whole process took 6-8 hours per day and results were issued in parallel to serology. The method was shown to detect all six HCV genotypes and a sensitivity of 500 IU/mL was achieved (95% hit rate). Until July 2005, 139,678 donations were tested and 315 (0.23%) were found reactive for HCV-RNA. Except for five false-positives, all 310 presented the corresponding antibody as well, so the yield of NAT-only donations was zero, presenting a specificity of 99.83%. Detection of a window period donation, in the population studied, will probably demand testing of a larger number of donations. International experience is showing a rate of 1:200,000 - 1:500,000 of isolated HCV-RNA reactive donations.

Optimization of the Tartrate-Resistant Acid Phosphatase Detection by Histochemical Method

According to the new KDIGO (Kidney Disease Improving Global Outcomes) guidelines, the term of renal osteodystrophy, should be used exclusively in reference to the invasive diagnosis of bone abnormalities. Due to the low sensitivity and specificity of biochemical serum markers of bone remodelling,the performance of bone biopsies is highly stimulated in dialysis patients and after kidney transplantation. The tartrate-resistant acid phosphatase (TRACP) is an iso-enzyme of the group of acid phosphatases, which is highly expressed by activated osteoclasts and macrophages. TRACP in osteoclasts is in intracytoplasmic vesicles that transport the products of bone matrix degradation. Being present in activated osteoclasts, the identification of this enzyme by histochemistry in undecalcified bone biopsies is an excellent method to quantify the resorption of bone. Since it is an enzymatic histochemical method for a thermolabile enzyme, the temperature at which it is performed is particularly relevant. This study aimed to determine the optimal temperature for identification of TRACP in activated osteoclasts in undecalcified bone biopsies embedded in methylmethacrylate. We selected 10 cases of undecalcified bone biopsies from hemodialysis patients with the diagnosis of secondary hyperparathyroidism. Sections of 5 μm were stained to identify TRACP at different incubation temperatures (37º, 45º, 60º, 70º and 80ºC) for 30 minutes. Activated osteoclasts stained red and trabecular bone (mineralized bone) was contrasted with toluidine blue. This approach also increased the visibility of the trabecular bone resorption areas (Howship lacunae). Unlike what is suggested in the literature and in several international protocols, we found that the best results were obtained with temperatures between 60ºC and 70ºC. For technical reasons and according to the results of the present study, we recommended that, for an incubation time of 30 minutes, the reaction should be carried out at 60ºC. As active osteoclasts are usually scarce in a bone section, the standardization of the histochemistry method is of great relevance, to optimize the identification of these cells and increase the accuracy of the histomosphometric results. Our results, allowing an increase in osteoclasts contrast, also support the use of semi-automatic histomorphometric measurements.

MOLECULAR IDENTIFICATION AND ANTIMICROBIAL RESISTANCE PATTERN OF SEVEN CLINICAL ISOLATES OF Nocardia spp. IN BRAZIL

Nocardia is a ubiquitous microorganism related to pyogranulomatous infection, which is difficult to treat in humans and animals. The occurrence of the disease is on the rise in many countries due to an increase in immunosuppressive diseases and treatments. This report of cases from Brazil presents the genotypic characterization and the antimicrobial susceptibility pattern using the disk-diffusion method and inhibitory minimal concentration with E-test® strips. In summary, this report focuses on infections in young adult men, of which three cases were cutaneous, two pulmonary, one neurological and one systemic. The pulmonary, neurological and systemic cases were attributed to immunosuppressive diseases or treatments. Sequencing analysis of the 16S rRNA segments (1491 bp) identified four isolates of Nocardia farcinica, two isolates of Nocardia nova and one isolate of Nocardia asiatica. N. farcinica was involved in two cutaneous, one systemic and other pulmonary cases; N. nova was involved in one neurological and one pulmonary case; and Nocardia asiatica in one cutaneous case. The disk-diffusion antimicrobial susceptibility test showed that the most effective antimicrobials were amikacin (100%), amoxicillin/clavulanate (100%), cephalexin (100%) and ceftiofur (100%), while isolates had presented most resistance to gentamicin (43%), sulfamethoxazole/trimethoprim (43%) and ampicillin (29%). However, on the inhibitory minimal concentration test (MIC test), only one of the four isolates of Nocardia farcinica was resistant to sulfamethoxazole/trimethoprim.

COMPARISON BETWEEN FOUR USUAL METHODS OF IDENTIFICATION OF Candida SPECIES

SUMMARY Infection by Candidaspp. is associated with high mortality rates, especially when treatment is not appropriate and/or not immediate. Therefore, it is necessary to correctly identify the genus and species of Candida. The aim of this study was to compare the identification of 89 samples of Candida spp. by the manual methods germ tube test, auxanogram and chromogenic medium in relation to the ID 32C automated method. The concordances between the methods in ascending order, measured by the Kappa index were: ID 32C with CHROMagar Candida(κ = 0.38), ID 32C with auxanogram (κ = 0.59) and ID 32C with germ tube (κ = 0.9). One of the species identified in this study was C. tropicalis,which demonstrated a sensitivity of 46.2%, a specificity of 95.2%, PPV of 80%, NPV of 81.1%, and an accuracy of 80.9% in tests performed with CHROMagar Candida;and a sensitivity of 76.9%, a specificity of 96.8%, PPV of 90.9%, NPV of 91%, and an accuracy of 91% in the auxanogram tests. Therefore, it is necessary to know the advantages and limitations of methods to choose the best combination between them for a fast and correct identification of Candidaspecies.

ISOLATION AND MOLECULAR IDENTIFICATION OF POTENTIALLY PATHOGENIC Escherichia coli AND Campylobacter jejuni IN FERAL PIGEONS FROM AN URBAN AREA IN THE CITY OF LIMA, PERU

SUMMARY Feral pigeons (Columbia livia) live in close contact with humans and other animals. They can transmit potentially pathogenic and zoonotic agents. The objective of this study was to isolate and detect strains of diarrheagenic Escherichia coli and Campylobacter jejuniof urban feral pigeons from an area of Lima, Peru. Fresh dropping samples from urban parks were collected for microbiological isolation of E. coli strains in selective agar, and Campylobacterby filtration method. Molecular identification of diarrheagenic pathotypes of E.coliand Campylobacter jejuni was performed by PCR. Twenty-two parks were sampled and 16 colonies of Campylobacter spp. were isolated. The 100% of isolates were identified as Campylobacter jejuni. Furthermore, 102 colonies of E. coli were isolated and the 5.88% resulted as Enteropathogenic (EPEC) type and 0.98% as Shiga toxin-producing E. coli (STEC). The urban feral pigeons of Lima in Peru can act as a reservoir or carriers of zoonotic potentially pathogenic enteric agents.

Immunocytochemical identification of leishmania and Trypanosoma cruzi amastigotes in situ with homologous and heterologous polyclonal antibodies

The unlabelled antibody peroxidase-antiperoxidase method was used to study the immunocytochemical properties of Leishmania and Trypanosoma cruzi amastigotes in situ after tissues had been submitted to different fixation procedures. Antisera were obtained from rabbits chronically infected with different strains of T. cruzi or immunized with L. mexicana amazonensis and L. braziliensis guyanensis, and were applied on 5 µm thick sections. T. cruzi antigens were well stained by the three anti-T. cruzi sera and the two anti-heis.hmama.sera at optimum dilution between 1:1,000 and 1:2,000, regardless the parasite strain. Differently, the leishmanial antigens were revealed by Leishmania sera only at low dilutions (between 1:60 -1:160), whereas the anti-T. cruzi sera, at these low dilutions, gave rather weak stainings. Although there is no clear explanation for this immunocytochemical "reverse-monodirectional" cross-reactivity between Leishmania and T. cruzi, the present results show that polyclonal antibodies agains Leishmania species, when used for immunocytochemical detection of these parasites in situ, react more strongly with T. cruzi amastigotes than with the homologous amastigotes.

Botnet cluster identification

Botnets are a group of computers infected with a specific sub-set of a malware family and controlled by one individual, called botmaster. This kind of networks are used not only, but also for virtual extorsion, spam campaigns and identity theft. They implement different types of evasion techniques that make it harder for one to group and detect botnet traffic. This thesis introduces one methodology, called CONDENSER, that outputs clusters through a self-organizing map and that identify domain names generated by an unknown pseudo-random seed that is known by the botnet herder(s). Aditionally DNS Crawler is proposed, this system saves historic DNS data for fast-flux and double fastflux detection, and is used to identify live C&Cs IPs used by real botnets. A program, called CHEWER, was developed to automate the calculation of the SVM parameters and features that better perform against the available domain names associated with DGAs. CONDENSER and DNS Crawler were developed with scalability in mind so the detection of fast-flux and double fast-flux networks become faster. We used a SVM for the DGA classififer, selecting a total of 11 attributes and achieving a Precision of 77,9% and a F-Measure of 83,2%. The feature selection method identified the 3 most significant attributes of the total set of attributes. For clustering, a Self-Organizing Map was used on a total of 81 attributes. The conclusions of this thesis were accepted in Botconf through a submited article. Botconf is known conferênce for research, mitigation and discovery of botnets tailled for the industry, where is presented current work and research. This conference is known for having security and anti-virus companies, law enforcement agencies and researchers.