Structured output SVM for remote sensing image classification

In the recent years, kernel methods have revealed very powerful tools in many application domains in general and in remote sensing image classification in particular. The special characteristics of remote sensing images (high dimension, few labeled samples and different noise sources) are efficiently dealt with kernel machines. In this paper, we propose the use of structured output learning to improve remote sensing image classification based on kernels. Structured output learning is concerned with the design of machine learning algorithms that not only implement input-output mapping, but also take into account the relations between output labels, thus generalizing unstructured kernel methods. We analyze the framework and introduce it to the remote sensing community. Output similarity is here encoded into SVM classifiers by modifying the model loss function and the kernel function either independently or jointly. Experiments on a very high resolution (VHR) image classification problem shows promising results and opens a wide field of research with structured output kernel methods.

Generalized descriptive set theory and classification theory

Descriptive set theory is mainly concerned with studying subsets of the space of all countable binary sequences. In this paper we study the generalization where countable is replaced by uncountable. We explore properties of generalized Baire and Cantor spaces, equivalence relations and their Borel reducibility. The study shows that the descriptive set theory looks very different in this generalized setting compared to the classical, countable case. We also draw the connection between the stability theoretic complexity of first-order theories and the descriptive set theoretic complexity of their isomorphism relations. Our results suggest that Borel reducibility on uncountable structures is a model theoretically natural way to compare the complexity of isomorphism relations.

Numerical Taxonomy of Old World Phlebotominae (Diptera: Psychodidae): 2. Restatement of Classification upon Subgeneric Morphological Characters

Numerical analyses (correspondence analysis, ascending hierarchical classification, and cladistics) were done with morphological characters of adult phlebotomine sand flies. The resulting classification largely confirms that of classical taxonomy for supra-specific groups from the Old World, though the positions of some groups are adjusted. The taxa Spelaeophlebotomus Theodor 1948, Idiophlebotomus Quate & Fairchild 1961, Australophlebotomus Theodor 1948 and Chinius Leng 1987 are notably distinct from other Old World groups, particularly from the genus Phlebotomus Rondani & Berté 1840. Spelaeomyia Theodor 1948 and, in particular, Parvidens Theodor & Mesghali 1964 are clearly separate from Sergentomyia França & Parrot 1920.

Textural classification of land cover using support vector machines: an empirical comparison with parametric, non parametric and hybrid classifiers in the Bolivian Amazon

Land cover classification is a key research field in remote sensing and land change science as thematic maps derived from remotely sensed data have become the basis for analyzing many socio-ecological issues. However, land cover classification remains a difficult task and it is especially challenging in heterogeneous tropical landscapes where nonetheless such maps are of great importance. The present study aims to establish an efficient classification approach to accurately map all broad land cover classes in a large, heterogeneous tropical area of Bolivia, as a basis for further studies (e.g., land cover-land use change). Specifically, we compare the performance of parametric (maximum likelihood), non-parametric (k-nearest neighbour and four different support vector machines - SVM), and hybrid classifiers, using both hard and soft (fuzzy) accuracy assessments. In addition, we test whether the inclusion of a textural index (homogeneity) in the classifications improves their performance. We classified Landsat imagery for two dates corresponding to dry and wet seasons and found that non-parametric, and particularly SVM classifiers, outperformed both parametric and hybrid classifiers. We also found that the use of the homogeneity index along with reflectance bands significantly increased the overall accuracy of all the classifications, but particularly of SVM algorithms. We observed that improvements in producer’s and user’s accuracies through the inclusion of the homogeneity index were different depending on land cover classes. Earlygrowth/degraded forests, pastures, grasslands and savanna were the classes most improved, especially with the SVM radial basis function and SVM sigmoid classifiers, though with both classifiers all land cover classes were mapped with producer’s and user’s accuracies of around 90%. Our approach seems very well suited to accurately map land cover in tropical regions, thus having the potential to contribute to conservation initiatives, climate change mitigation schemes such as REDD+, and rural development policies.

Human Plasma-derived Polymeric IgA and IgM Antibodies Associate with Secretory Component to Yield Biologically Active Secretory-like Antibodies.

Immunotherapy with monoclonal and polyclonal immunoglobulin is successfully applied to improve many clinical conditions, including infection, autoimmune diseases, or immunodeficiency. Most immunoglobulin products, recombinant or plasma-derived, are based on IgG antibodies, whereas to date, the use of IgA for therapeutic application has remained anecdotal. In particular, purification or production of large quantities of secretory IgA (SIgA) for potential mucosal application has not been achieved. In this work, we sought to investigate whether polymeric IgA (pIgA) recovered from human plasma is able to associate with secretory component (SC) to generate SIgA-like molecules. We found that ∼15% of plasma pIgA carried J chain and displayed selective SC binding capacity either in a mixture with monomeric IgA (mIgA) or after purification. The recombinant SC associated covalently in a 1:1 stoichiometry with pIgA and with similar efficacy as colostrum-derived SC. In comparison with pIgA, the association with SC delayed degradation of SIgA by intestinal proteases. Similar results were obtained with plasma-derived IgM. In vitro, plasma-derived IgA and SIgA neutralized Shigella flexneri used as a model pathogen, resulting in a delay of bacteria-induced damage targeted to polarized Caco-2 cell monolayers. The sum of these novel data demonstrates that association of plasma-derived IgA or IgM with recombinant/colostrum-derived SC is feasible and yields SIgA- and SIgM-like molecules with similar biochemical and functional characteristics as mucosa-derived immunoglobulins.

Polygodial, the fungitoxic component from the Brazilian medicinal plant Polygonum punctatum

Polygonum punctatum (Polygonaceae) is an herb known in some regions of Brazil as "erva-de-bicho" and is used to treat intestinal disorders. The dichloromethane extract of the aerial parts of this plant showed strong activity in a bioautographic assay with the fungus Cladosporium sphaerospermum. The bioassay-guided chemical fractionation of this extract afforded the sesquiterpene dialdehyde polygodial as the active constituent. The presence of this compound with antibiotic, anti-inflammatory and anti-hyperalgesic properties in "erva-de-bicho" may account for the effects attributed by folk medicine to this plant species.

International consensus conference on PFAPA syndrome: evaluation of a new set of classification criteria

INTRODUCTION: PFAPA syndrome is characterized by periodic fever, associated with pharyngitis, cervical adenitis and/or aphthous stomatitis and belongs to the auto-inflammatory diseases. Diagnostic criteria are based on clinical features and the exclusion of other periodic fever syndromes. An analysis of a large cohort of patients has shown weaknesses for these criteria and there is a lack of international consensus. An International Conference was held in Morges in November 2008 to propose a new set of classification criteria based on a consensus among experts in the field.OBJECTIVE: We aimed to verify the applicability of the new set of classification criteria.PATIENTS & METHODS: 80 patients diagnosed with PFAPA syndrome from 3 centers (Genoa, Lausanne and Geneva) for pediatric rheumatology were included in the study. A detailed description of the clinical and laboratory features was obtained. The new classification criteria and the actual diagnostic criteria were applied to the patients.RESULTS: Only 40/80 patients (50%) fulfilled all criteria of the new classification. 31 patients were excluded because they didn't meet one of the 7 diagnostic criteria, 7 because of 2 criteria, and one because of 3 criteria. When we applied the current criteria to the same patients, 11/80 patients (13.7%) needed to be excluded. 8/80 patients (10%) were excluded from both sets. Exclusion was related only to some of the criteria. Number of patients for each not fulfilled criterion (new set of criteria/actual criteria): age (1/6), symptoms between episodes (2/2), delayed growth (4/1), main symptoms (21/0), periodicity, length of fever, interval between episodes, and length of disease (20/0). The application of some of the new criteria was not easy, as they were both very restrictive and needed precise information from the patients.CONCLUSION: Our work has shown that the new set of classification criteria can be applied to patients suspected for PFAPA syndrome, but it seems to be more restrictive than the actual diagnostic criteria. A further work of validation needs to be done in order to determine if this new set of classification criteria allow a good discrimination between PFAPA patients and other causes of recurrent fever syndromes.

Classification of human astrocytic gliomas on the basis of gene expression: a correlated group of genes with angiogenic activity emerges as a strong predictor of subtypes.

The development of targeted treatment strategies adapted to individual patients requires identification of the different tumor classes according to their biology and prognosis. We focus here on the molecular aspects underlying these differences, in terms of sets of genes that control pathogenesis of the different subtypes of astrocytic glioma. By performing cDNA-array analysis of 53 patient biopsies, comprising low-grade astrocytoma, secondary glioblastoma (respective recurrent high-grade tumors), and newly diagnosed primary glioblastoma, we demonstrate that human gliomas can be differentiated according to their gene expression. We found that low-grade astrocytoma have the most specific and similar expression profiles, whereas primary glioblastoma exhibit much larger variation between tumors. Secondary glioblastoma display features of both other groups. We identified several sets of genes with relatively highly correlated expression within groups that: (a). can be associated with specific biological functions; and (b). effectively differentiate tumor class. One prominent gene cluster discriminating primary versus nonprimary glioblastoma comprises mostly genes involved in angiogenesis, including VEGF fms-related tyrosine kinase 1 but also IGFBP2, that has not yet been directly linked to angiogenesis. In situ hybridization demonstrating coexpression of IGFBP2 and VEGF in pseudopalisading cells surrounding tumor necrosis provided further evidence for a possible involvement of IGFBP2 in angiogenesis. The separating groups of genes were found by the unsupervised coupled two-way clustering method, and their classification power was validated by a supervised construction of a nearly perfect glioma classifier.

Immunogenicity test of tetanus component in adsorbed vaccines by toxin binding inhibition test

Samples from 20 lots of diphtheria-tetanus (adult use dT) vaccine and from 20 lots of diphtheria-tetanus-pertussis (DTP) vaccine were used to standardize and validate the in vitro toxin binding inhibition (ToBI) test for the immunogenicity test of the tetanus component. The levels of tetanus antitoxin obtained by ToBI test were compared to those obtained using the toxin neutralization (TN) test in mice routinely employed to perform the quality control of the tetanus component in adsorbed vaccines. The results ranged from 1.8 to 3.5 IU/ml for dT and 2 to 4 IU/ml for DTP by ToBI test and 1.4 to 3 IU/ml for dT and 1.8 to 3.5 IU/ml for DTP by TN in mice. These results were significantly correlated. From this study, it is concluded that the ToBI test is an alternative to the in vivo neutralization procedure in the immunogenicity test of the tetanus component in adsorbed vaccines. A substantial refinement and a reduction in use of animals can be achieved.

Detection of antibodies to the 97 kDa component of Toxoplasma gondii in samples of human serum

This study was carried out to investigate the immune response against 97 kDa (p97) molecular marker of Toxoplasma gondii that has been characterized as a cytosolic protein and a component of the excreted-secreted antigens from this parasite. A total of 60 serum samples from patients were analyzed by enzime-linked immunosorbent assay and Western blot for toxoplasmosis. These samples were organized in three groups, based on clinical symptoms and results of serological tests. Group I: 20 samples reactive to IgG and IgM (acute phase); group II: 20 non-reactive samples (control group); and group III: 20 samples reactive only to IgG (chronic phase). Western blot was performed with total antigenic extracts or with excreted and secreted antigen from T. gondii to identify the fraction correspondent to p97. It was observed that this cytosolic component from T. gondii stimulates the immunologic system to produce both IgM and IgG antibodies in the beginning of the acute infection and IgG throughout the chronic stage of the asymptomatic toxoplasmosis.