Independent component analysis applied to raman spectra for classification of in vitro human coronary arteries


Autoria(s): SILVEIRA JR., Landulfo; PAULA JR., Alderico Rodrigues De; PASQUALUCCI, Carlos Augusto; PACHECO, Marcos Tadeu T.
Contribuinte(s)

UNIVERSIDADE DE SÃO PAULO

Data(s)

19/10/2012

19/10/2012

2008

Resumo

Optical diagnostic methods, such as near-infrared Raman spectroscopy allow quantification and evaluation of human affecting diseases, which could be useful in identifying and diagnosing atherosclerosis in coronary arteries. The goal of the present work is to apply Independent Component Analysis (ICA) for data reduction and feature extraction of Raman spectra and to perform the Mahalanobis distance for group classification according to histopathology, obtaining feasible diagnostic information to detect atheromatous plaque. An 830nm Ti:sapphire laser pumped by an argon laser provides near-infrared excitation. A spectrograph disperses light scattered from arterial tissues over a liquid-nitrogen cooled CCD to detect the Raman spectra. A total of 111 spectra from arterial fragments were utilized.

Identificador

INSTRUMENTATION SCIENCE & TECHNOLOGY, v.36, n.2, p.134-145, 2008

1073-9149

http://producao.usp.br/handle/BDPI/22718

10.1080/10739140701850845

http://dx.doi.org/10.1080/10739140701850845

Idioma(s)

eng

Publicador

TAYLOR & FRANCIS INC

Relação

Instrumentation Science & Technology

Direitos

restrictedAccess

Copyright TAYLOR & FRANCIS INC

Palavras-Chave #Raman spectroscopy #Independent Component Analysis (ICA) #Mahalanobis distance #human coronary artery #MAHALANOBIS DISTANCE #SOURCE SEPARATION #ANALYSIS ICA #ATHEROSCLEROSIS #SPECTROSCOPY #SIGNALS #Chemistry, Analytical #Instruments & Instrumentation
Tipo

article

original article

publishedVersion