Spectrochemical differentiation in endometriosis based on infrared spectroscopy advanced data fusion and multivariate...

Endometriosis is a common benign gynecological condition characterized by the growth of endometrial gland and stroma located outside the uterine cavity, which the current approaches for its detection are invasive and expensive, limiting their clinical utility. There is a need for cost-effective and minimally invasive approaches to facilitate the diagnosis of this disease. Attenuated total reflection Fourier-transform infrared and near infrared spectroscopy combined with multivariate classification were applied as a new tool to analyze blood plasma samples from women with endometriosis (n = 41) and healthy individuals (n = 34). In addition, the use of advanced data fusion strategies and multivariate analysis techniques improved the classification models and facilitated diagnostics segregation of both sample categories in a fast and non-destructive way, generating high levels of accuracy, sensitivity and specificity. 2D correlation analysis revealed strong positive correlations between the spectrochemical biomarkers identified in both IR regions. To the best of our knowledge, this is the first study demonstrating the efficacy of a new tool for fast and non-invasive diagnosis of endometriosis using blood plasma samples analyzed with IR spectroscopy combined with multivariate classification.