Abstract
Lung cancer diagnosis in early stages could be of paramount interest since patients may be treated opportunely decreasing the high death rate caused by this disease. A biomarker may describe abnormalities in the human being and may be correlated with a specific illness. Currently, no single biomarker reported has proved to be sufficiently specific and sensitive for lung cancer, thus the search is an open research issue. In this document a set of fourteen biomarkers were evaluated jointly for lung cancer detection, nevertheless, interpreting the information from these biomarkers is a quite complex task and powerful computational tools are required for proper data analysis. Thus an Artificial Neural Network was trained with a set of lung cancer biomarkers. Principal Component Analysis allowed reducing the biomarkers initial vector from fourteen to seven proteins. The Artificial Neural Network performed satisfactorily classifying correctly 60 out of 64 individuals. ANN trained with seven biomarkers -MMP-1, MMP- 9, Cyfra 21-1, CRP, CEA, YKL-40, CA-125- yielded an increase in sensitivity of approximately 20%, i.e., 98.97%, compared with that of the best single biomarker, Cyfra 21-1 (sensitivity 78.9%). The corresponding specificity was 80%. ANN significantly improved the sensitivity of biomarkers, therefore ANN offers a promising auxiliary tool in diagnosis of lung cancer.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2013 Springer Berlin Heidelberg
About this paper
Cite this paper
Flores, J.M. et al. (2013). Artificial Neural Network-Based Serum Biomarkers Analysis Improves Sensitivity in the Diagnosis of Lung Cancer. In: Folgueras Méndez, J., et al. V Latin American Congress on Biomedical Engineering CLAIB 2011 May 16-21, 2011, Habana, Cuba. IFMBE Proceedings, vol 33. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-21198-0_224
Download citation
DOI: https://doi.org/10.1007/978-3-642-21198-0_224
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-21197-3
Online ISBN: 978-3-642-21198-0
eBook Packages: EngineeringEngineering (R0)