Multivariate Statistical Analysis of Raman Spectra of Human Skin for Chronic Heart Failure Diagnosis
Yulia A. Khristoforova, Samara University, Russia
Lyudmila A. Bratchenko, Samara University, Russia
Petr A. Lebedev, Samara State Medical University, Russia
Maria A. Skuratova, Samara City Clinical Hospital №1 named N.I. Pirogova, Russia
Elena A. Lebedeva, Samara City Clinical Hospital №1 named N.I. Pirogova, Russia
Ivan A. Bratchenko, Samara University, Russia
Abstract
In this study, Raman spectroscopy in combination with chemometrics is studied as an alternative noninvasive fast method to identify chronic heart failure cases. Optical analysis is based on the analysis of spectral features changes associated with the biochemical composition changes of skin tissues. Portable spectroscopy setup with 785 nm excitation wavelength was used to register skin Raman spectral features. In this in vivo study, 127 patients and 57 healthy volunteers were involved for measurements of skin spectral features by Raman spectroscopy. The spectral data were analyzed with a projection on the latent structures and discriminant analysis. Classification of 202 skin spectra of patients with chronic heart failure and 90 skin spectra of healthy volunteers was performed with 0.888 ROC AUC for 10-fold cross validated algorithm. The performance of proposed classifier was estimated on the new test set that equaled to the 0.917 ROC AUC to identify chronic heart failure cases.
File with abstract
Speaker
Yulia Khristoforova
Samara National Research University
Russian Federation
Discussion
Ask question
