Lecture : “Radiogenomics” (so called, imaging genomics) is focused on defining relationships between cancer imaging features and molecular phenotypes. Radiogenomics can be mainly used to create imaging biomarkers that can identify the genomics of malignant disease without the use of a biopsy. This approach has been successfully applied in glioblastoma, breast cancer, and lung cancer based on MRI and CT images. In terms of liver cancer, Segal E et al. found that combinations of 28 imaging traits in CT could reconstruct 78% of the global gene expression profiles which revealed cell proliferation, liver synthetic function, and patient prognosis. They also proved that radiogenomic venous invasion (RVI), which is a CT biomarker of microvascular invasion derived from a 91-gene expression signature, accurately predicted microvascular invasion in hepatocellular carcinoma (HCC) specimens and was associated with postoperative outcomes. 18F-FDG PET is a diagnostic tool on the basis of differential glucose metabolism between tumor and normal tissue. PET positive HCC may be associated with poorly differentiated disease and unfavorable outcomes after treatment. PET-based radiogenomics studies showed that 14 quantitative PET imaging features describing FDG uptake were correlated with gene expression for 8 single genes and coexpressed gene clusters (metagenes), and 4 of these single genes (LY6E, RNF149, MCM6, and FAP) were associated with survival of lung cancer patients. In a breast cancer study, presence of the FDG signature was associated with MYC gene copy gain, and increased MYC transcript levels and upregulation of metabolic MYC target genes in cancer patients. We aimed to create a radiogenomic map linking metabolic PET images and genotranscriptomic profiles generated by whole exome sequencing and RNA-seq for patients with resected HCC. Our analysis showed significant associations between PET hypermetabolism and overexpression of genes related to mTOR signaling and cell cycle regulation in HCC.