IκB-ζ controls the constitutive NF-κB target gene network and survival of ABC DLBCL.

Constitutive activation of the nuclear factor-κ B (NF-κB) pathway is a hallmark of the activated B-cell-like (ABC) subtype of diffuse large B-cell lymphoma (DLBCL). Recurrent mutations of NF-κB regulators that cause constitutive activity of this oncogenic pathway have been identified. However, it remains unclear how specific target genes are regulated. We identified the atypical nuclear IκB protein IκB-ζ to be upregulated in ABC compared with germinal center B-cell-like (GCB) DLBCL primary patient samples. Knockdown of IκB-ζ by RNA interference was toxic to ABC but not to GCB DLBCL cell lines. Gene expression profiling after IκB-ζ knockdown demonstrated a significant downregulation of a large number of known NF-κB target genes, indicating an essential role of IκB-ζ in regulating a specific set of NF-κB target genes. To further investigate how IκB-ζ mediates NF-κB activity, we performed immunoprecipitations and detected a physical interaction of IκB-ζ with both p50 and p52 NF-κB subunits, indicating that IκB-ζ interacts with components of both the canonical and the noncanonical NF-κB pathway in ABC DLBCL. Collectively, our data demonstrate that IκB-ζ is essential for nuclear NF-κB activity in ABC DLBCL, and thus might represent a promising molecular target for future therapies.

Red meat and cancer risk in a network of case-control studies focusing on cooking practices.

BACKGROUND: Consumption of red meat has been related to increased risk of several cancers. Cooking methods could modify the magnitude of this association, as production of chemicals depends on the temperature and duration of cooking. METHODS: We analyzed data from a network of case-control studies conducted in Italy and Switzerland between 1991 and 2009. The studies included 1465 oral and pharyngeal, 198 nasopharyngeal, 851 laryngeal, 505 esophageal, 230 stomach, 1463 colon, 927 rectal, 326 pancreatic, 3034 breast, 454 endometrial, 1031 ovarian, 1294 prostate and 767 renal cancer cases. Controls included 11 656 patients admitted for acute, non-neoplastic conditions. Odds ratios (ORs) and confidence intervals (CIs) were estimated by multiple logistic regression models, adjusted for known confounding factors. RESULTS: Daily intake of red meat was significantly associated with the risk of cancer of the oral cavity and pharynx (OR for increase of 50 g/day = 1.38; 95% CI: 1.26-1.52), nasopharynx (OR = 1.29; 95% CI: 1.04-1.60), larynx (OR = 1.46; 95% CI: 1.30-1.64), esophagus (OR = 1.46; 95% CI: 1.23-1.72), colon (OR = 1.17; 95% CI: 1.08-1.26), rectum (OR = 1.22; 95% CI:1.11-1.33), pancreas (OR = 1.51; 95% CI: 1.25-1.82), breast (OR = 1.12; 95% CI: 1.04-1.19), endometrium (OR = 1.30; 95% CI: 1.10-1.55) and ovary (OR = 1.29; 95% CI: 1.16-1.43). Fried meat was associated with a higher risk of cancer of oral cavity and pharynx (OR = 2.80; 95% CI: 2.02-3.89) and esophagus (OR = 4.52; 95% CI: 2.50-8.18). Risk of prostate cancer increased for meat cooked by roasting/grilling (OR = 1.31; 95% CI: 1.12-1.54). No heterogeneity according to cooking methods emerged for other cancers. Nonetheless, significant associations with boiled/stewed meat also emerged for cancer of the nasopharynx (OR = 1.97; 95% CI: 1.30-3.00) and stomach (OR = 1.86; 95% CI: 1.20-2.87). CONCLUSIONS: Our analysis confirmed red meat consumption as a risk factor for several cancer sites, with a limited impact of cooking methods. These findings, thus, call for a limitation of its consumption in populations of Western countries.

Outcome and patterns of failure in testicular lymphoma: a multicenter Rare Cancer Network study.

PURPOSE: To assess the outcome and patterns of failure in patients with testicular lymphoma treated by chemotherapy (CT) and/or radiation therapy (RT). METHODS AND MATERIALS: Data from a series of 36 adult patients with Ann Arbor Stage I (n = 21), II (n = 9), III (n = 3), or IV (n = 3) primary testicular lymphoma, consecutively treated between 1980 and 1999, were collected in a retrospective multicenter study by the Rare Cancer Network. Median age was 64 years (range: 21-91 years). Full staging workup (chest X-ray, testicular ultrasound, abdominal ultrasound, and/or thoracoabdominal computer tomography, bone marrow assessment, full blood count, lactate dehydrogenase, and cerebrospinal fluid evaluation) was completed in 18 (50%) patients. All but one patient underwent orchidectomy, and spermatic cord infiltration was found in 9 patients. Most patients (n = 29) had CT, consisting in most cases of cyclophosphamide, doxorubicin, vincristine, and prednisolone (CHOP) with (n = 17) or without intrathecal CT. External RT was delivered to scrotum alone (n = 12) or testicular, iliac, and para-aortic regions (n = 8). The median RT dose was 31 Gy (range: 20-44 Gy) in a median of 17 fractions (10-24), using a median of 1.8 Gy (range: 1.5-2.5 Gy) per fraction. The median follow-up period was 42 months (range: 6-138 months). RESULTS: After a median period of 11 months (range: 1-76 months), 14 patients presented lymphoma progression, mostly in the central nervous system (CNS) (n = 8). Among the 17 patients who received intrathecal CT, 4 had a CNS relapse (p = NS). No testicular, iliac, or para-aortic relapse was observed in patients receiving RT to these regions. The 5-year overall, lymphoma-specific, and disease-free survival was 47%, 66%, and 43%, respectively. In univariate analyses, statistically significant factors favorably influencing the outcome were early-stage and combined modality treatment. Neither RT technique nor total dose influenced the outcome. Multivariate analysis revealed that the most favorable independent factors predicting the outcome were younger age, early-stage disease, and combined modality treatment. CONCLUSIONS: In this multicenter retrospective study, CNS was found to be the principal site of relapse, and no extra-CNS lymphoma progression was observed in the irradiated volumes. More effective CNS prophylaxis, including combined modalities, should be prospectively explored in this uncommon site of extranodal lymphoma.

Primary mucosa-associated lymphoid tissue lymphoma of the salivary glands: a multicenter Rare Cancer Network study.

Purpose: Involvement of salivary glands with mucosa-associated lymphoid tissue (MALT) lymphoma is rare. This retrospective study was performed to assess the clinical profile, treatment outcome, and prognostic factors of MALT lymphoma of the salivary glands.Methods and Materials: Thirteen member centers of the Rare Cancer Network from 10 countries participated, providing data on 63 patients. The median age was 58 years; 47 patients were female and 16 were male. The parotid glands were involved in 49 cases, submandibular in 15, and minor glands in 3. Multiple glands were involved in 9 patients. Staging was as follows: IE in 34, IIE in 12, IIIE in 2, and IV in 15 patients.Results: Surgery (S) alone was performed in 9, radiotherapy (RI) alone in 8, and chemotherapy (CT) alone in 4 patients. Forty-one patients received combined modality treatment (S + RT in 23, S + CT in 8, RT + CT in 4, and all three modalities in 6 patients). No active treatment was given in one case. After initial treatment there was no tumor in 57 patients and residual tumor in 5. Tumor progression was observed in 23 (36.5%) (local in 1, other salivary glands in 10, lymph nodes in 11, and elsewhere in 6). Five patients died of disease progression and the other 5 of other causes. The 5-year disease-free survival, disease-specific survival, and overall survival were 54.4%, 93.2%, and 81.7%, respectively. Factors influencing disease-free survival were use of RI, stage, and residual tumor (p < 0.01). Factors influencing disease-specific survival were stage, recurrence, and residual tumor (p < 0.01).Conclusions: To our knowledge, this report represents the largest series of MALT lymphomas of the salivary glands published to date. This disease may involve all salivary glands either initially or subsequently in 30% of patients. Recurrences may occur in up to 35% of patients at 5 years; however, survival is not affected. Radiotherapy is the only treatment modality that improves disease-free survival. (C) 2012 Elsevier Inc.

A protein interaction atlas for the nuclear receptors: properties and quality of a hub-based dimerisation network.

BACKGROUND: The nuclear receptors are a large family of eukaryotic transcription factors that constitute major pharmacological targets. They exert their combinatorial control through homotypic heterodimerisation. Elucidation of this dimerisation network is vital in order to understand the complex dynamics and potential cross-talk involved. RESULTS: Phylogeny, protein-protein interactions, protein-DNA interactions and gene expression data have been integrated to provide a comprehensive and up-to-date description of the topology and properties of the nuclear receptor interaction network in humans. We discriminate between DNA-binding and non-DNA-binding dimers, and provide a comprehensive interaction map, that identifies potential cross-talk between the various pathways of nuclear receptors. CONCLUSION: We infer that the topology of this network is hub-based, and much more connected than previously thought. The hub-based topology of the network and the wide tissue expression pattern of NRs create a highly competitive environment for the common heterodimerising partners. Furthermore, a significant number of negative feedback loops is present, with the hub protein SHP [NR0B2] playing a major role. We also compare the evolution, topology and properties of the nuclear receptor network with the hub-based dimerisation network of the bHLH transcription factors in order to identify both unique themes and ubiquitous properties in gene regulation. In terms of methodology, we conclude that such a comprehensive picture can only be assembled by semi-automated text-mining, manual curation and integration of data from various sources.

Outcome and Prognostic Factors in Adenosquamous Carcinoma of the Head and Neck: A Multicenter Rare Cancer Network Study

Purpose/Objective(s): Adenosquamous carcinoma (AC) of the head and neck is a distinct entity first described in 1968. Its natural history is more aggressive than squamous cell carcinoma but this is based on very small series reported in the literature. The goal of this study was to assess the clinical profile, outcome, patterns of failure and prognostic factors in patients with AC of the head and neck treated by radiation therapy (RT) with or without chemotherapy (CT).Materials/Methods: Data from 18 patients with Stage I (n = 3), II (n = 1), III (n = 4), or IVa (n = 10) AC, treated between 1989 and 2009, were collected in a retrospective multicenter Rare Cancer Network study. Median age was 60 years (range, 48 - 73 years). Fourteen patients were male and 4 female. Risk factors, including perineural invasion, lymphangitis, vascular invasion, positive margins, were present in 83% of the patients. Tumor sites included oral cavity in 4, oropharynx in 4, hypopharynx in2, larynx in 2, salivary glands in 2, nasal vestibule in 2, nasopharynx in 1, and maxillary sinus in 1 patient. Surgery (S) was performed in all but 5 patients. S alone was performed in only 1 patient, and definitive RT alone in 3 patients. Fourteen patients received combined modality treatment (S+RT in 10, RT+CT in 2, and all of the three modalities in 2 patients). Median RT dose to the primary and to the nodes was 66 Gy (range, 50 - 72 Gy) and 53 Gy (range, 44 - 66 Gy), respectively (1.8 - 2.0 Gy/fr., 5 fr./ week). In 4 patients, the planning treatment volume included the primary tumor site only. Seven patients were treated with 2D RT, 7 with 3D conformal RT, and 2 with intensity-modulated RT.Results: After a median follow-up period of 38 months (range, 9 - 62 months), 8 patients developed distant metastases (lung, bone, mediastinum, and liver), 6 presented nodal recurrences, and only 4 had a local relapse at the primary site (all in-field recurrences). At last follow-up, 6 patients were alive without disease, 1 alive with disease, 9 died from progressive disease, and 2 died from intercurrent disease. The 3-year and median overall survival, disease-free survival (DFS) and locoregional control rates were 52% (95% confidence interval [CI]: 28 - 76%) and 39 months, 36% (95% CI: 13 - 49%) and 12 months, and 54% (95% CI: 26 - 82%) and 40 months, respectively. In multivariate analysis (Cox model), DFS was negatively influenced by the presence of extracapsular extension (p = 0.02) and advanced stage (IV versus I-III, p = 0.003).Conclusions: Overall prognosis of locoregionally advanced AC remains poor, and distant metastases and nodal relapse occur in almost half of the cases. However, local control is relatively good, and early stage AC patients had prolonged DFS when treated with combined modality treatment.

Radioimmunotherapy confers long-term survival to lymphoma patients with acceptable toxicity: registry analysis by the international radioimmunotherapy network.

The Radioimmunotherapy Network (RIT-N) is a Web-based, international registry collecting long-term observational data about radioimmunotherapy-treated patients with malignant lymphoma outside randomized clinical studies. The RIT-N collects unbiased data on treatment indications, disease stages, patients' conditions, lymphoma subtypes, and hematologic side effects of radioimmunotherapy treatment. Methods: RIT-N is located at the University of Gottingen, Germany, and collected data from 14 countries. Data were entered by investigators into a Web-based central database managed by an independent clinical research organization. Results: Patients (1,075) were enrolled from December 2006 until November 2009, and 467 patients with an observation time of at least 12 mo were included in the following analysis. Diagnoses were as follows: 58% follicular lymphoma and 42% other B-cell lymphomas. The mean overall survival was 28 mo for follicular lymphoma and 26 mo for other lymphoma subtypes. Hematotoxicity was mild for hemoglobin (World Health Organization grade II), with a median nadir of 10 g/dL, but severe (World Health Organization grade III) for platelets and leukocytes, with a median nadir of 7,000/mu L and 2.2/mu L, respectively. Conclusion: Clinical usage of radioimmunotherapy differs from the labeled indications and can be assessed by this registry, enabling analyses of outcome and toxicity data beyond clinical trials. This analysis proves that radioimmunotherapy in follicular lymphoma and other lymphoma subtypes is a safe and efficient treatment option.

Analysis, modelling and classification of geospatial data using machine learning

The research considers the problem of spatial data classification using machine learning algorithms: probabilistic neural networks (PNN) and support vector machines (SVM). As a benchmark model simple k-nearest neighbor algorithm is considered. PNN is a neural network reformulation of well known nonparametric principles of probability density modeling using kernel density estimator and Bayesian optimal or maximum a posteriori decision rules. PNN is well suited to problems where not only predictions but also quantification of accuracy and integration of prior information are necessary. An important property of PNN is that they can be easily used in decision support systems dealing with problems of automatic classification. Support vector machine is an implementation of the principles of statistical learning theory for the classification tasks. Recently they were successfully applied for different environmental topics: classification of soil types and hydro-geological units, optimization of monitoring networks, susceptibility mapping of natural hazards. In the present paper both simulated and real data case studies (low and high dimensional) are considered. The main attention is paid to the detection and learning of spatial patterns by the algorithms applied.

Validation of network communicability metrics for the analysis of brain structural networks.

Computational network analysis provides new methods to analyze the brain's structural organization based on diffusion imaging tractography data. Networks are characterized by global and local metrics that have recently given promising insights into diagnosis and the further understanding of psychiatric and neurologic disorders. Most of these metrics are based on the idea that information in a network flows along the shortest paths. In contrast to this notion, communicability is a broader measure of connectivity which assumes that information could flow along all possible paths between two nodes. In our work, the features of network metrics related to communicability were explored for the first time in the healthy structural brain network. In addition, the sensitivity of such metrics was analysed using simulated lesions to specific nodes and network connections. Results showed advantages of communicability over conventional metrics in detecting densely connected nodes as well as subsets of nodes vulnerable to lesions. In addition, communicability centrality was shown to be widely affected by the lesions and the changes were negatively correlated with the distance from lesion site. In summary, our analysis suggests that communicability metrics that may provide an insight into the integrative properties of the structural brain network and that these metrics may be useful for the analysis of brain networks in the presence of lesions. Nevertheless, the interpretation of communicability is not straightforward; hence these metrics should be used as a supplement to the more standard connectivity network metrics.

Analysis, modelling and visualisation of spatiotemporal data for urban studies

The proportion of population living in or around cites is more important than ever. Urban sprawl and car dependence have taken over the pedestrian-friendly compact city. Environmental problems like air pollution, land waste or noise, and health problems are the result of this still continuing process. The urban planners have to find solutions to these complex problems, and at the same time insure the economic performance of the city and its surroundings. At the same time, an increasing quantity of socio-economic and environmental data is acquired. In order to get a better understanding of the processes and phenomena taking place in the complex urban environment, these data should be analysed. Numerous methods for modelling and simulating such a system exist and are still under development and can be exploited by the urban geographers for improving our understanding of the urban metabolism. Modern and innovative visualisation techniques help in communicating the results of such models and simulations. This thesis covers several methods for analysis, modelling, simulation and visualisation of problems related to urban geography. The analysis of high dimensional socio-economic data using artificial neural network techniques, especially self-organising maps, is showed using two examples at different scales. The problem of spatiotemporal modelling and data representation is treated and some possible solutions are shown. The simulation of urban dynamics and more specifically the traffic due to commuting to work is illustrated using multi-agent micro-simulation techniques. A section on visualisation methods presents cartograms for transforming the geographic space into a feature space, and the distance circle map, a centre-based map representation particularly useful for urban agglomerations. Some issues on the importance of scale in urban analysis and clustering of urban phenomena are exposed. A new approach on how to define urban areas at different scales is developed, and the link with percolation theory established. Fractal statistics, especially the lacunarity measure, and scale laws are used for characterising urban clusters. In a last section, the population evolution is modelled using a model close to the well-established gravity model. The work covers quite a wide range of methods useful in urban geography. Methods should still be developed further and at the same time find their way into the daily work and decision process of urban planners. La part de personnes vivant dans une région urbaine est plus élevé que jamais et continue à croître. L'étalement urbain et la dépendance automobile ont supplanté la ville compacte adaptée aux piétons. La pollution de l'air, le gaspillage du sol, le bruit, et des problèmes de santé pour les habitants en sont la conséquence. Les urbanistes doivent trouver, ensemble avec toute la société, des solutions à ces problèmes complexes. En même temps, il faut assurer la performance économique de la ville et de sa région. Actuellement, une quantité grandissante de données socio-économiques et environnementales est récoltée. Pour mieux comprendre les processus et phénomènes du système complexe "ville", ces données doivent être traitées et analysées. Des nombreuses méthodes pour modéliser et simuler un tel système existent et sont continuellement en développement. Elles peuvent être exploitées par le géographe urbain pour améliorer sa connaissance du métabolisme urbain. Des techniques modernes et innovatrices de visualisation aident dans la communication des résultats de tels modèles et simulations. Cette thèse décrit plusieurs méthodes permettant d'analyser, de modéliser, de simuler et de visualiser des phénomènes urbains. L'analyse de données socio-économiques à très haute dimension à l'aide de réseaux de neurones artificiels, notamment des cartes auto-organisatrices, est montré à travers deux exemples aux échelles différentes. Le problème de modélisation spatio-temporelle et de représentation des données est discuté et quelques ébauches de solutions esquissées. La simulation de la dynamique urbaine, et plus spécifiquement du trafic automobile engendré par les pendulaires est illustrée à l'aide d'une simulation multi-agents. Une section sur les méthodes de visualisation montre des cartes en anamorphoses permettant de transformer l'espace géographique en espace fonctionnel. Un autre type de carte, les cartes circulaires, est présenté. Ce type de carte est particulièrement utile pour les agglomérations urbaines. Quelques questions liées à l'importance de l'échelle dans l'analyse urbaine sont également discutées. Une nouvelle approche pour définir des clusters urbains à des échelles différentes est développée, et le lien avec la théorie de la percolation est établi. Des statistiques fractales, notamment la lacunarité, sont utilisées pour caractériser ces clusters urbains. L'évolution de la population est modélisée à l'aide d'un modèle proche du modèle gravitaire bien connu. Le travail couvre une large panoplie de méthodes utiles en géographie urbaine. Toutefois, il est toujours nécessaire de développer plus loin ces méthodes et en même temps, elles doivent trouver leur chemin dans la vie quotidienne des urbanistes et planificateurs.

Citrus fruit and cancer risk in a network of case-control studies.

BACKGROUND: Citrus fruit has shown a favorable effect against various cancers. To better understand their role in cancer risk, we analyzed data from a series of case-control studies conducted in Italy and Switzerland. PATIENTS AND METHODS: The studies included 955 patients with oral and pharyngeal cancer, 395 with esophageal, 999 with stomach, 3,634 with large bowel, 527 with laryngeal, 2,900 with breast, 454 with endometrial, 1,031 with ovarian, 1,294 with prostate, and 767 with renal cell cancer. All cancers were incident and histologically confirmed. Controls were admitted to the same network of hospitals for acute, nonneoplastic conditions. Odds ratios (OR) were estimated by multiple logistic regression models, including terms for major identified confounding factors for each cancer site, and energy intake. RESULTS: The ORs for the highest versus lowest category of citrus fruit consumption were 0.47 (95% confidence interval, CI, 0.36-0.61) for oral and pharyngeal, 0.42 (95% CI, 0.25-0.70) for esophageal, 0.69 (95% CI, 0.52-0.92) for stomach, 0.82 (95% CI, 0.72-0.93) for colorectal, and 0.55 (95% CI, 0.37-0.83) for laryngeal cancer. No consistent association was found with breast, endometrial, ovarian, prostate, and renal cell cancer. CONCLUSIONS: Our findings indicate that citrus fruit has a protective role against cancers of the digestive and upper respiratory tract.

Automatic Mapping and Classification of Spatial Environmental Data

The paper deals with the development and application of the generic methodology for automatic processing (mapping and classification) of environmental data. General Regression Neural Network (GRNN) is considered in detail and is proposed as an efficient tool to solve the problem of spatial data mapping (regression). The Probabilistic Neural Network (PNN) is considered as an automatic tool for spatial classifications. The automatic tuning of isotropic and anisotropic GRNN/PNN models using cross-validation procedure is presented. Results are compared with the k-Nearest-Neighbours (k-NN) interpolation algorithm using independent validation data set. Real case studies are based on decision-oriented mapping and classification of radioactively contaminated territories.

Multitask learning of environmental spatial data

The present research deals with an application of artificial neural networks for multitask learning from spatial environmental data. The real case study (sediments contamination of Geneva Lake) consists of 8 pollutants. There are different relationships between these variables, from linear correlations to strong nonlinear dependencies. The main idea is to construct a subsets of pollutants which can be efficiently modeled together within the multitask framework. The proposed two-step approach is based on: 1) the criterion of nonlinear predictability of each variable ?k? by analyzing all possible models composed from the rest of the variables by using a General Regression Neural Network (GRNN) as a model; 2) a multitask learning of the best model using multilayer perceptron and spatial predictions. The results of the study are analyzed using both machine learning and geostatistical tools.

Diffusion Spectrum Imaging after stroke shows structural changes in the contra-lateral motor network correlating with functional recovery

There is growing interest in understanding the role of the non-injured contra-lateral hemisphere in stroke recovery. In the experimental field, histological evidence has been reported that structural changes occur in the contra-lateral connectivity and circuits during stroke recovery. In humans, some recent imaging studies indicated that contra-lateral sub-cortical pathways and functional and structural cortical networks are remodeling, after stroke. Structural changes in the contra-lateral networks, however, have never been correlated to clinical recovery in patients. To determine the importance of the contra-lateral structural changes in post-stroke recovery, we selected a population of patients with motor deficits after stroke affecting the motor cortex and/or sub-cortical motor white matter. We explored i) the presence of Generalized Fractional Anisotropy (GFA) changes indicating structural alterations in the motor network of patientsâeuro? contra-lateral hemisphere as well as their longitudinal evolution ii) the correlation of GFA changes with patientsâeuro? clinical scores, stroke size and demographics data iii) and a predictive model.

Data-driven clustering : new methods and applications

Abstract : This work is concerned with the development and application of novel unsupervised learning methods, having in mind two target applications: the analysis of forensic case data and the classification of remote sensing images. First, a method based on a symbolic optimization of the inter-sample distance measure is proposed to improve the flexibility of spectral clustering algorithms, and applied to the problem of forensic case data. This distance is optimized using a loss function related to the preservation of neighborhood structure between the input space and the space of principal components, and solutions are found using genetic programming. Results are compared to a variety of state-of--the-art clustering algorithms. Subsequently, a new large-scale clustering method based on a joint optimization of feature extraction and classification is proposed and applied to various databases, including two hyperspectral remote sensing images. The algorithm makes uses of a functional model (e.g., a neural network) for clustering which is trained by stochastic gradient descent. Results indicate that such a technique can easily scale to huge databases, can avoid the so-called out-of-sample problem, and can compete with or even outperform existing clustering algorithms on both artificial data and real remote sensing images. This is verified on small databases as well as very large problems. Résumé : Ce travail de recherche porte sur le développement et l'application de méthodes d'apprentissage dites non supervisées. Les applications visées par ces méthodes sont l'analyse de données forensiques et la classification d'images hyperspectrales en télédétection. Dans un premier temps, une méthodologie de classification non supervisée fondée sur l'optimisation symbolique d'une mesure de distance inter-échantillons est proposée. Cette mesure est obtenue en optimisant une fonction de coût reliée à la préservation de la structure de voisinage d'un point entre l'espace des variables initiales et l'espace des composantes principales. Cette méthode est appliquée à l'analyse de données forensiques et comparée à un éventail de méthodes déjà existantes. En second lieu, une méthode fondée sur une optimisation conjointe des tâches de sélection de variables et de classification est implémentée dans un réseau de neurones et appliquée à diverses bases de données, dont deux images hyperspectrales. Le réseau de neurones est entraîné à l'aide d'un algorithme de gradient stochastique, ce qui rend cette technique applicable à des images de très haute résolution. Les résultats de l'application de cette dernière montrent que l'utilisation d'une telle technique permet de classifier de très grandes bases de données sans difficulté et donne des résultats avantageusement comparables aux méthodes existantes.