Premium
Discrimination of tumor from normal tissues in a mouse model of breast cancer using CARS spectroscopy combined with PC‐DFA methodology
Author(s) -
Huang Xi,
Yuan Ye,
Bielecki Timothy A.,
Mohapatra Bhopal C.,
Luan Haitao,
SilvaLopez Edibaldo,
West William W.,
Band Vimla,
Lu Yongfeng,
Band Hamid,
Zhang Tian C.
Publication year - 2017
Publication title -
journal of raman spectroscopy
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.748
H-Index - 110
eISSN - 1097-4555
pISSN - 0377-0486
DOI - 10.1002/jrs.5201
Subject(s) - principal component analysis , breast cancer , discriminant function analysis , linear discriminant analysis , cancer , pathology , artificial intelligence , medicine , computer science , mathematics , statistics
Objective methodologies to discriminate tumor from normal tissue in biopsies and resection specimens are of great interest as complementary approaches to existing pathological diagnosis of tumors. In the present study, coherent anti‐Stokes Raman scattering (CARS) spectroscopy was applied as an approach to discriminate resected tumor from normal mammary tissue in murine mammary tumor virus‐Wnt‐1 transgenic mouse model of breast cancer. Due to the dense CH molecular vibration in the range from 2500 to 3100 cm −1 , the classification was performed by using principal component‐discriminant function analysis to discriminate tumor from the normal tissue. A total of 240 training and 40 testing CARS spectra were acquired. The overall accuracy of CARS, based on cross‐validation and external validation method, was 98% and 95%, respectively. The present study demonstrates a diagnostic method with a 1‐s spectral acquirement rate, using a CARS spectroscopic technique. Our results suggest that CARS combined with the principal component‐discriminant function analysis is a potentially useful tool for identification and classification of breast cancer tissues. Copyright © 2017 John Wiley & Sons, Ltd.