Document Detail


Study of support vector machine and serum surface-enhanced Raman spectroscopy for noninvasive esophageal cancer detection.
MedLine Citation:
PMID:  23389685     Owner:  NLM     Status:  In-Data-Review    
Abstract/OtherAbstract:
ABSTRACT. The ability of combining serum surface-enhanced Raman spectroscopy (SERS) with support vector machine (SVM) for improving classification esophageal cancer patients from normal volunteers is investigated. Two groups of serum SERS spectra based on silver nanoparticles (AgNPs) are obtained: one group from patients with pathologically confirmed esophageal cancer (n=30) and the other group from healthy volunteers (n=31). Principal components analysis (PCA), conventional SVM (C-SVM) and conventional SVM combination with PCA (PCA-SVM) methods are implemented to classify the same spectral dataset. Results show that a diagnostic accuracy of 77.0% is acquired for PCA technique, while diagnostic accuracies of 83.6% and 85.2% are obtained for C-SVM and PCA-SVM methods based on radial basis functions (RBF) models. The results prove that RBF SVM models are superior to PCA algorithm in classification serum SERS spectra. The study demonstrates that serum SERS in combination with SVM technique has great potential to provide an effective and accurate diagnostic schema for noninvasive detection of esophageal cancer.
Authors:
Shao-Xin Li; Qiu-Yao Zeng; Lin-Fang Li; Yan-Jiao Zhang; Ming-Ming Wan; Zhi-Ming Liu; Hong-Lian Xiong; Zhou-Yi Guo; Song-Hao Liu
Related Documents :
23297055 - The prognosis of gastric cancer patients with marginally elevated carcinoembryonic anti...
1903585 - Quantitation of energy expenditure by infrared thermography.
20149555 - Could stroma contribute to field cancerization?
23749485 - Association between nqo1 c609t polymorphism and bladder cancer susceptibility: a system...
24957465 - Reading between the lines; understanding drug response in the post genomic era.
1372305 - The dream in terminal illness: a jungian formulation.
Publication Detail:
Type:  Journal Article    
Journal Detail:
Title:  Journal of biomedical optics     Volume:  18     ISSN:  1560-2281     ISO Abbreviation:  J Biomed Opt     Publication Date:  2013 Feb 
Date Detail:
Created Date:  2013-02-07     Completed Date:  -     Revised Date:  -    
Medline Journal Info:
Nlm Unique ID:  9605853     Medline TA:  J Biomed Opt     Country:  United States    
Other Details:
Languages:  eng     Pagination:  27008     Citation Subset:  IM    
Affiliation:
South China Normal University, School of Information and Optoelectronic Science and Engineering, Guangzhou 510631, ChinabGuangdong Medical College, School of Information Engineering, Dongguan 523808, China.
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms
Descriptor/Qualifier:

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine


Previous Document:  Quantifying the cerebral metabolic rate of oxygen by combining diffuse correlation spectroscopy and ...
Next Document:  Associations of Depressive and Anxiety Disorders with Periodontal Disease Prevalence in Young Adults...