Chronic Obstructive Pulmonary Disease (COPD) is a preventable, treatable, and slowly progressive disease, whose course is aggravated by a periodic worsening of symptoms and lung function lasting for several days. The need to implement personalized treatment, where the characteristics of the patients together with disease information will be used to select the best treatment option, has boosted the research for identifying COPD phenotypes. This asks for addressing data clustering and feature selection, but both have shown some weaknesses when applied to this aim. To overcome such limitations, in this work we simultaneously select the discriminative descriptors and cluster the data. Our idea stems from observing that such two tasks are strictly interrelated each other, motivating us for using a method where, iteration by iteration, feature selection influences the clustering step, and viceversa. As a results we discover five phenotypes that, contrary to the traditional classes defined by the Global Initiative for Chronic Obstructive Lung Disease, seem to be prone to specific outcomes. This is of particular importance from a clinical point of view because it will allow a more tailored management of the disease.

Discovering COPD phenotyping via simultaneous feature selection and clustering

Merone M;Pedone C;Antonelli Incalzi R;Iannello G;Soda P
2019-01-01

Abstract

Chronic Obstructive Pulmonary Disease (COPD) is a preventable, treatable, and slowly progressive disease, whose course is aggravated by a periodic worsening of symptoms and lung function lasting for several days. The need to implement personalized treatment, where the characteristics of the patients together with disease information will be used to select the best treatment option, has boosted the research for identifying COPD phenotypes. This asks for addressing data clustering and feature selection, but both have shown some weaknesses when applied to this aim. To overcome such limitations, in this work we simultaneously select the discriminative descriptors and cluster the data. Our idea stems from observing that such two tasks are strictly interrelated each other, motivating us for using a method where, iteration by iteration, feature selection influences the clustering step, and viceversa. As a results we discover five phenotypes that, contrary to the traditional classes defined by the Global Initiative for Chronic Obstructive Lung Disease, seem to be prone to specific outcomes. This is of particular importance from a clinical point of view because it will allow a more tailored management of the disease.
2019
978-153865488-0
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.12610/16614
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 1
  • ???jsp.display-item.citation.isi??? ND
social impact