Analysis of THz and Raman spectra of glioma patients biofluids by machine learning methods
Denis A. Vrazhnov1, Olga P. CHerkasova2, Yuri V. Kistenev3; 1 V.E. Zuev Institute of Atmospheric Optics SB RAS, Tomsk, Russia; 2Institute of Automation and Electrometry, SB RAS, Novosibirsk, Russia; 3Tomsk State University, Tomsk, Russia
Abstract
Application of modern machine learning methods to the analysis of glioma is presented. Blood serum of small animals, blood plasma and liquor of human patients was studied by the THz and Raman spectroscopy. Methods for informative biomarkers detection was evaluated and compared with magnetic resonance spectroscopy data. Problems of analysis highly dimensional but low samples number biodata is discussed. Examples of machine learning pipelines with dimensionality reduction based on PCA, t-SNE and classification models constructed by XGBoost, SVM, Random forests are presented.
The research was carried out with the support of a grant under the Decree of the Government of the Russian Federation No. 220 of 09 April 2010 (Agreement No. 075-15-2021-615 of 04 June 2021)
File with abstract
Speaker
Denis A. Vrazhnov
IAO SB RAS
Russia
Discussion
Ask question
