Radiomics and radiogenomics of glioblastoma: theoretical foundations and clinical application possibilities. Literature review

Glioblastoma is a neuroepithelial malignant brain tumor with an aggressive clinical course and adverse prognosis. The most common symptoms are headache, seizures, neurological deficits, cognitive and personality disorders. On MRI, glioblastomas appear as a formation with fuzzy contours, heterogeneous structure, with zones of uneven accumulation of contrast medium surrounding the hypointensive necrotic center of the tumor. The median overall survival is 15 months after complex treatment, which dictates the need to develop a personalized approach in the diagnosis and treatment of glioblastomas. The basic principle of radiomics involves extracting a large number of quantitative features from medical images using computer algorithms; radiogenomics is a tool for assessing tumor subtype, mutational status and intra-tumor heterogeneity, reflecting the relationship with progression, survival and response to treatment. The use of radiomics and radiogenomics in glioblastomas creates the potential for predicting survival, differential diagnosis of gliomas and other tumors, determiningtumor Grade, detecting mutations and amplifications, differential diagnosis of pseudoprogression and tumor progression, predicting the response to chemoradiotherapy.