Radiomics is an evolving field in which extraction and screening a large amount of image features could be used to estab－lish predictive models with clinical information through artificial intelligence technology and statistical analysis，thereby improving the diagnostic accuracy，predicting treatment response and prognostication，potentially allowing personalized cancer therapy. As a promis－ing new field，radiomics has broad application prospects. This review will focus on the overview，current status，process and challenges in the clinical application of PET radiomics in non-small cell lung cancer.