Buongiorno a Tutti,
Si informa che lunedi’ 12 settembre presso il Dipartimento di Matematica dell’Universita’ di Genova, aula 715, si terranno i due seguenti seminari.
Cordiali saluti, Eva Riccomagno
ore 15.:30 Manuele Leonelli, Instituto de Matemática, Universidade Federal do Rio De Janeiro, UFRJ https://sites.google.com/site/manueleleonelli https://sites.google.com/site/manueleleonelli TITLE: Bayesian semiparametric multivariate models for extreme exceedances ABSTRACT: Interest on extremal events generally involves the joint study of many concomitant variables, as for instance wave height and surge. We build on previous work which specifically accounted for marginal exceedances over a high, unknown threshold, by combining it with flexible families of copulae. This approach allows for the detection of specific patterns of dependence be them extremal or not. Attention is also devoted to the ascertainment of asymptotic independence, where standard multivariate extreme value theory is not applicable. Estimation of higher quantiles and other quantities of interest is performed both marginally and conditionally via MCMC algorithms. Our approach is evaluated through a series of simulations and is applied to real data sets.
ore 16:30 Alessio Signori, Dipartimento di Scienze della Salute, Universita’ degli Studi di Genova TITLE: Longitudinal trajectories of EDSS in primary progressive multiple sclerosis patients A latent class approach ABSTRACT: Background. Over the last decades several natural history studies on primary progressive MS (PPMS) patients were reported from international registries. In this population a consistent heterogeneity was observed in the rate of disability accumulation, as time to reach the milestone of Expanded Disability Status Scale (EDSS) 6 ranged between 7 and 14 years from onset. Objectives. To identify subgroups of PPMS patients with similar longitudinal trajectories of EDSS over time. Methods. All PPMS patients collected within the MSBase international registry, who had their first EDSS assessment within 5 years from onset were included in the analysis. Longitudinal EDSS scores were modelled by a latent class mixed model (LCMM), using a nonlinear function of time from onset. LCMM is an advanced statistical approach that models heterogeneity between patients by classifying them into unobserved groups (latent classes) showing similar characteristics. Results. A total of 853 PPMS (51.7% females) from 24 countries with a mean age at onset of 42.4 yrs (SD: 10.8 yrs), a median baseline EDSS of 4 (IQR: 2.5-5.5) and 2.4 yrs of disease duration (SD: 1.5 yrs) were included. LCMM detected 3 different subgroups of patients with a mild (n=143 ;16.8%), moderate (n=378; 44.3%) or severe (n=332; 38.9%) disability trajectory. Median time to EDSS 4 was 14, 5 and 3.7 years respectively, for the 3 groups. The probability of reaching EDSS 6 at 10 years was 0%, 46.5% and 83.1% respectively. Using this modelling approach it is possible to predict the future disease course of a subject with PPMS using early EDSS assessments. Using only 1 year of EDSS monitoring 73% of patients are correctly classified in their disability trajectory group (mild, moderate or severe); after 3 years this proportion increases to 87% and after 5 years it reaches 92%. Conclusions. Using long-term observations and an LCMM modelling approach, it is possible to build a dynamic model, to predict the future disability trajectory of newly diagnosed PPMS patients. In the design of future clinical trials in PPMS, using time to reach disability milestones as the primary endpoint, the existence of heterogeneous classes of patients should be considered.
________________________________________________________ Prof. Eva Riccomagno Dipartimento di Matematica - Universita` degli Studi di Genova Via Dodecaneso, 35 - 16146 Genova - ITALIA Tel: +39 - 010 - 353 6938 Fax: +39 - 010 - 353 6960 www.dima.unige.it/~riccomag