Research and Progress of Artificial Intelligence in the Imaging Genomics of Lung Cancer

Abstract

Imaging genomics based on artificial intelligence technology has shown great potential in personalized treatment and precision medicine, providing important theoretical basis and technical support for the intelligent diagnosis and prognosis prediction of lung cancer. In order to facilitate the more effective exploration and wider application of diagnosis and prognosis of lung cancer, the relevant research progress is described in detail, expecting that such review will be beneficial to the researchers in this field. First, it introduces the role and application of artificial intelligence in image genomics, and the main research direction of artificial intelligence technology in lung cancer imaging genomics, which are reviewed from lung cancer gene phenotype identification, image gene bidirectional correlation analysis, and prognosis prediction. In addition, the advantages and disadvantages of radiomics and deep learning algorithm are evaluated in each direction, as well as the facing problems and challenges. Second, it summarizes the main challenges facing intelligent imaging genomics at this stage and presents future research directions. Although imaging genomics based on artificial intelligence has made some achievements in lung cancer diagnosis, survival recurrence prediction, efficacy response evaluation, and understanding tumor biological mechanism, it also needs to combining the actual clinical needs, establish a unified, standard, perfect computer-assisted clinical program further explore interpretability, reproducibility, and universal verification for the application of artificial intelligence in lung cancer imaging genomics, so as to provide guarantee for the intelligent diagnosis treatment and evaluation of lung cancer.

Keywords

• artificial intelligence
• imaging genomics
• lung cancer
• gene phenotype identification
• bidirectional correlation analysis
• prognosis prediction