Noninvasive glioblastoma diagnosis using spectral methods and machine learning

Noninvasive glioblastoma diagnosis can be achieved by analyzing blood by Terahertz, Infrared and Raman spectroscopy. The model of xenotransplantation of the U87 human glioblastoma cells into immunodeficient mice was used. The most informative frequencies, separating glioblastoma's and healthy groups were identified by machine learning methods.