In and ex vivo breast disease study by Raman spectroscopy

Publication: Research - peer-reviewJournal article – Annual report year: 2011

  • Author: Raniero, L.

    Universidade do Vale do Paraíba, Brazil, Institute of Research and Development, IP&D

  • Author: Canevari, R. A.

    Universidade do Vale do Paraíba, Brazil, Institute of Research and Development, IP&D

  • Author: Ramalho, L. N. Z.

    University of São Paulo

  • Author: Ramalho, F. S.

    University of São Paulo

  • Author: dos Santos, E. A. P.

    Universidade do Vale do Paraíba, Brazil, Institute of Research and Development, IP&D

  • Author: Bitar, R. A.

    University of São Paulo

  • Author: Jalkanen, Karl J.

    Department of Physics, Technical University of Denmark, Denmark

  • Author: Martinho, H. S.

    Universidade Federal do ABC, Brazil, Centro de Ciências Naturais e Humanas

  • Author: Martin, A. A.

    Universidade do Vale do Paraíba, Brazil, Institute of Research and Development, IP&D

View graph of relations

In this work, Raman spectra in the 900-1,800 cm(-1) wavenumber region of in vivo and ex vivo breast tissues of both healthy mice (normal) and mice with induced mammary gland tumors (abnormal) were measured. In the case of the in vivo tissues, the Raman spectra were collected for both transcutaneous (with skin) and skin-removed tissues. To identify the spectral differences between normal and cancer breast tissue, the paired t-test was carried out for each wavenumber using the whole spectral range from both groups. Quadratic discriminate analysis based on principal component analysis (PCA) was also used to determine and evaluate differences in the Raman spectra for the various samples as a basis for diagnostic purposes. The differences in the Raman spectra of the samples were due to biochemical changes at the molecular, cellular and tissue levels. The sensitivity and specificity of the classification scheme based on the differences in the Raman spectra obtained by PCA were evaluated using the receiver operating characteristic (ROC) curve. The in vivo transcutaneous normal and abnormal tissues were correctly classified based on their measured Raman spectra with a discriminant proportion of 73%, while the in vivo skin-removed normal and abnormal tissues were correctly classified again based on their measured Raman spectra with a discriminant proportion of 86%. This result reveals a strong influence due to the skin of the breast, which decreased the specificity by 11%. Finally, the results from ex vivo measurements gave the highest specificity and sensitivity: 96 and 97%, respectively, as well as a largest percentage for correct discrimination: 94%. Now that the important bands have been experimentally determined in this and other works, what remains is for first principles molecular-level simulations to determine whether the changes are simply due to conformational changes, due to aggregation, due to changes in the environment, or complex interactions of all of the above.
Original languageEnglish
JournalTheoretical Chemistry Accounts
Publication date2011
Volume130
Issue4-6
Pages1239-1247
ISSN1432-881X
DOIs
StatePublished
CitationsWeb of Science® Times Cited: 7

Keywords

  • Principal component analysis, Raman spectroscopy, Quadratic discriminant analysis, Cancer diagnostic, Receiver operator characteristic curve, Optical biopsy, Cancer detection, Breast cancer
Download as:
Download as PDF
Select render style:
APAAuthorCBEHarvardMLAStandardVancouverShortLong
PDF
Download as HTML
Select render style:
APAAuthorCBEHarvardMLAStandardVancouverShortLong
HTML
Download as Word
Select render style:
APAAuthorCBEHarvardMLAStandardVancouverShortLong
Word

ID: 6386919