Detection of lymph node metastases in cholangiocanma by fourier transform infrared spectroscopy

Background: The fast development of Fourier transform infrared (FTIR) spectroscopy has proposed a novel avenue for distinguishing cancerous tissue from normal one. The aim of this research is to investigate the possibility of employing FTIR spectroscopy with support vector machine (SVM) classification as a new spectral discriminant method for classifying metastatic and non-metastatic lymph nodes. Methods: A total of 63 lymph nodes, from cholangiocarcinoma patients who underwent surgery for cholangiocarcinomaectomy, were obtained and then measured by FTIR spectroscopy. The average FTIR spectrum of metastatic and non-metastatic lymph nodes was made. Standard normal variate (SNV) method was exploited to reduce scatter effect, and support vector machine (SVM) classification model was utilized to discriminate spectrum of metastatic from non-metastatic lymph nodes. Leaveone-out cross validation (LOOCV) was used to evaluate SVM model effects. Results: 20 metastatic and 43 nonmetastatic lymph nodes were pathologically diagnosed. Two parameters, including I1743 (related to Lipid) and I1080 (related to Nucleic acid), are founded statistically different between metastatic and non-metastatic lymph nodes. The average FTIR spectra of metastatic group was pronouncedly different from non-metastatic group. The accuracy, sensitivity, and specificity of the SVM model were 88.9%, 75.0%, 95.3%, respectively. Conclusion: A novel approach to differentiate metastatic and non-metastatic lymph nodes in cholangiocarcinoma by employing the FTIR spectroscopy combined with SVM model, is established and demonstrated, and could be applied in cholangiocarcinoma clinical stage value in future clinical practice.