Radiomics is an emerging field in quantitative imaging that uses advanced imaging features to objectively and quantitatively describe tumor phenotypes. This review discusses the applications and challenges of radiomics, highlighting its potential in personalized medicine. Radiomic features have shown promise in predicting treatment outcomes, differentiating benign from malignant tumors, assessing cancer genetics, and staging tumors. However, the quantification of radiomic features is influenced by various technical factors, including acquisition modes, reconstruction parameters, image discretization schemes, and respiratory motion. The reproducibility of radiomic features and the need for proper study design, such as avoiding feature selection bias and ensuring adequate sample sizes, are also discussed. Despite these challenges, radiomics shows great potential in capturing important phenotypic information for personalized therapy. Standardization and harmonization of imaging acquisition and feature computation are crucial for consistent findings across institutions.Radiomics is an emerging field in quantitative imaging that uses advanced imaging features to objectively and quantitatively describe tumor phenotypes. This review discusses the applications and challenges of radiomics, highlighting its potential in personalized medicine. Radiomic features have shown promise in predicting treatment outcomes, differentiating benign from malignant tumors, assessing cancer genetics, and staging tumors. However, the quantification of radiomic features is influenced by various technical factors, including acquisition modes, reconstruction parameters, image discretization schemes, and respiratory motion. The reproducibility of radiomic features and the need for proper study design, such as avoiding feature selection bias and ensuring adequate sample sizes, are also discussed. Despite these challenges, radiomics shows great potential in capturing important phenotypic information for personalized therapy. Standardization and harmonization of imaging acquisition and feature computation are crucial for consistent findings across institutions.