Advances in artificial intelligence-based radiogenomics for lung cancer precision medicine

Artificial intelligence-based radiogenomics has emerged as a promising approach for precision medicine in lung cancer. By integrating medical imaging, genomics, and clinical data, radiogenomics enables non-invasive prediction of key oncogenic driver mutations, exploration of associations between imaging features and gene expression, and development of prognostic models in lung cancer management. Machine learning and deep learning techniques have been applied to predict the mutation status of genes such as epidermal growth factor receptor and Kirsten rat sarcoma viral oncogene homolog, which are crucial for personalized treatment strategies. Radiogenomic studies have identified significant correlations between radiomic features and gene clusters, providing insights into tumor heterogeneity and biological pathways. Moreover, radiogenomics has shown potential in predicting treatment responses, recurrence, and overall survival in lung cancer patients. However, challenges remain in standardization, comprehensive validation, model interpretability, ethnic diversity, and the construction of multi-omics databases. With the advancement of artificial intelligence and the expansion of multimodal databases, future research should focus on solving these challenges to improve the clinical value and generalizability of radiogenomic models, thus playing a greater role in personalized medicine for cancer.