Classification of Squamous Cell Carcinoma of the Oral Cavity Using Wavelet Analysis and BP-Chaos Networks

66 samples from the human oral mucosa tissue were measured by confocal Raman microspectroscopy. A preprocessed algorithm based on wavelet analysis was used to reduce noise and eliminate background of Raman spectra firstly. Then the integrated areas of four normalised wavenumber regions 1004, 1156, 1360 1587 and 1660 cm^-1 were carried out for discrimination of the normal and malignant oral mucosa tissue samples. Lastly, by means of combining the improved BP algorithm and chaos optimization algorithm, an embedded BP-Chaos algorithm was proposed and successfully used as a classification method to identify whether the tissue is normal or not. This could result in a new diagnostic method, which would assist the early diagnosis of squamous cell carcinoma of the oral cavity.