DocumentCode :
2095571
Title :
A Two-Step Approach for Feature Selection and Classifier Ensemble Construction in Computer-Aided Diagnosis
Author :
Lee, Michael C. ; Boroczky, Lilla ; Sungur-Stasik, Kivilcim ; Cann, Aaron D. ; Borczuk, Alain C. ; Kawut, Steven M. ; Powell, Charles A.
Author_Institution :
Philips Res. North America, Briarcliff Manor, NY
fYear :
2008
fDate :
17-19 June 2008
Firstpage :
548
Lastpage :
553
Abstract :
Accurate classification methods are critical in computer-aided diagnosis and other clinical decision support systems. Previous research has studied methods for combining genetic algorithms for feature selection with ensemble classifier systems in an effort to increase classification accuracy. We propose a two-step approach that first uses genetic algorithms to reduce the number of features used to characterize the data, then applies the random subspace method on the remaining features to create a set of diverse but high performing classifiers. These classifiers are combined using ensemble learning techniques to yield a final classification. We demonstrate this approach for computer-aided diagnosis of solitary pulmonary nodules from CT scans, in which the proposed method outperforms several previously described methods.
Keywords :
decision support systems; genetic algorithms; medical diagnostic computing; pattern classification; CT scans; classifier ensemble construction; clinical decision support systems; computer-aided diagnosis; ensemble classifier systems; feature selection; genetic algorithms; random subspace method; solitary pulmonary nodules; two-step approach; Biological cells; Cancer; Computed tomography; Computer aided diagnosis; Genetic algorithms; Lungs; Machine learning; Performance evaluation; USA Councils; Voting; classifier ensemble; computer-aided diagnosis; feature selection; genetic algorithm; lung cancer;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer-Based Medical Systems, 2008. CBMS '08. 21st IEEE International Symposium on
Conference_Location :
Jyvaskyla
ISSN :
1063-7125
Print_ISBN :
978-0-7695-3165-6
Type :
conf
DOI :
10.1109/CBMS.2008.68
Filename :
4562055
Link To Document :
بازگشت