Le vendredi  26 août 2022, Madame Halehsadat NEKOEE ZAHRAEI subira l'examen en vue de l'obtention du grade académique de Docteur en Sciences de la santé publique sous la direction de Madame Anne-Françoise DONNEAU et Monsieur Renaud LOUIS

Cette épreuve consistera en la défense publique d'un mémoire intitulé :

"Contribution to cluster analysis in chronic obstructive airway diseases"

Asthma and chronic obstructive pulmonary disease are widely accepted as complex and heterogeneous disorders with different clinical phenotypes. Understanding the phenotype of each patient is the first step toward effective personalized management and treatment. There are various clinical approaches to determine these phenotypes, however, the common statistical approach is cluster analysis. Cluster analysis of multidimensional variables and objects with missing values are unavoidable challenges in this thesis. Therefore, this thesis is a comprehensive study on cluster analysis, including two main parts: statistical methodologies, and their application to the clinical datasets, with the goal of classifying incomplete multidimensional clinical datasets. The first statistical contribution of this thesis is a proposed framework for handling missing values using multiple imputation and multi-dimensional data in cluster analysis which contains a new method based on mixture multivariate multinomial model(4M). In addition, the proposed framework was evaluated with parallel frameworks and methods in extensive scenarios using simulated datasets with known clustering results under different missingness and overlapping rates. Eventually, the proposed framework was applied to COPD, eosinophilic and also, non-eosinophilic datasets and classified them into homogenous clusters which mainly differentiate by demographics, atopic status, functional and inflammatory features. 

Le Jury sera composé de :

Bernard VRIJENS (Président), Sylvie STREEL (Secrétaire), Anne-Françoise DONNEAU, Renaud LOUIS, Catherine MOERMANS, Nicolas SAUVAGEOT (Johnson & Johnson, Statistic.&decision sce group), Florence SCHLEICH, Olivier VANDENPLAS (UCL)