Machine learning in analysis of blood Raman spectra
O. Cherkasova1,2, A. Mankova3, D. Vrazhnov4,5, Y. Kistenev5,6
1Institute of Laser Physics, Siberian Branch of the RAS, Novosibirsk, Russia
2Institute on Laser and Information Technologies - Branch of the Federal Scientific Research Centre "Crystallography and Photonics" of RAS, Shatura, Moscow Region, Russia
3Faculty of Physics, Lomonosov Moscow State University, Moscow, Russia,
4V.E. Zuev Institute of Atmospheric optics, Siberian Branch of the RAS, Tomsk, Russia
5Tomsk State University, Tomsk, Russia
6Siberian State Medical University, Tomsk, Russia
Abstract
Machine learning methods were used to identify the most informative frequencies associated with cancer molecular markers. Raman spectra of blood plasma were studied in the dynamics of the experimental cancer. We used a DXR Raman Microscope (Thermo Scientific), excitation wavelengths of 532 nm, range 80–3200 cm–1. Two tasks were considered in the framework of the analysis of Raman spectra using machine learning methods. First, the difference between the control group and the comparison group was studied with the informative feature selection that has the greatest impact on group separability. Second, the dynamics of changes in Raman spectra for cancer groups with an increase in the duration of tumor development was considered. The question of the influence of the background subtraction procedure on the choice of informative features by the method of principal components analysis (PCA), Random Forests (RF), and Support Vector Machines (SVM) was investigated separately.
This work was supported by the RFBR (grant # 19-52-55004), the Ministry of Science and Higher Education of the Russian Federation within the State assignment FSRC "Crystallography and Photonics" RAS, by the Interdisciplinary Scientific and Educational School of Moscow University “Photonic and Quantum Technologies. Digital Medicine”. 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
Olga Cherkasova
1Institute of Laser Physics, Siberian Branch, Russian Academy of Sciences, Novosibirsk,Russia 2Institute on Laser and Information Technologies - Branch of the Federal Scientific Research Centre "Crystallography and Photonics" of RAS, Shatura, Moscow Reg
Russia
Discussion
Ask question
