Journal of Innovative Optical Health Sciences, Volume. 9, Issue 5, 1650017(2016)

In vivo subsite classification and diagnosis of oral cancers using Raman spectroscopy

Aditi Sahu1... Atul Deshmukh1, Arti R. Hole1, Pankaj Chaturvedi2 and C. Murali Krishna1,* |Show fewer author(s)
Author Affiliations
  • 1Chilakapati Laboratory, ACTREC, TMC Kharghar, Navi Mumbai 410210, India
  • 2Head and Neck Surgical Oncology, Tata Memorial Hospital, Dr. E. Borges Road, Parel, Mumbai, Maharashtra- 400012, India
  • show less

    Oral cancers suffer from poor disease-free survival rates due to delayed diagnosis. Noninvasive, rapid, objective approaches as adjuncts to visual inspection can help in better management of oral cancers. Raman spectroscopy (RS) has shown potential in identification of oral premalignant and malignant conditions and also in the detection of early cancer changes like cancer-field-effects (CFE) at buccal mucosa subsite. Anatomic differences between different oral subsites have also been reported using RS. In this study, anatomical differences between subsites and their possible influence on healthy vs pathological classification were evaluated on 85 oral cancer and 72 healthy subjects. Spectra were acquired from buccal mucosa, lip and tongue in healthy, contralateral (internal healthy control), premalignant and cancer conditions using fiber-optic Raman spectrometer. Mean spectra indicate predominance of lipids in healthy buccal mucosa, contribution of both lipids and proteins in lip while major dominance of protein in tongue spectra. From healthy to tumor, changes in protein secondary-structure, DNA and heme-related features were observed. Principal component linear discriminant analysis (PC-LDA) followed by leave-one-out-crossvalidation (LOOCV) was used for data analysis. Findings indicate buccal mucosa and tongue are distinct entities, while lip misclassifies with both these subsites. Additionally, the diagnostic algorithm for individual subsites gave improved classification efficiencies with respect to the pooled subsites model. However, as the pooled subsites model yielded 98% specificity and 100% sensitivity, this model may be more useful for preliminary screening applications. Large-scale validation studies are a pre-requisite before envisaging future clinical applications.

    Tools

    Get Citation

    Copy Citation Text

    Aditi Sahu, Atul Deshmukh, Arti R. Hole, Pankaj Chaturvedi, C. Murali Krishna. In vivo subsite classification and diagnosis of oral cancers using Raman spectroscopy[J]. Journal of Innovative Optical Health Sciences, 2016, 9(5): 1650017

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Received: Aug. 21, 2015

    Accepted: Nov. 10, 2015

    Published Online: Dec. 27, 2018

    The Author Email: Krishna C. Murali (mchilakapati@actrec.gov.in)

    DOI:10.1142/s1793545816500176

    Topics