With the formation of an aging society, cataract caused by age has become a public common concern about health problem all over the world. Surgery by intraocular lens replacement is the only effective mean to treat cataract. The most important thing to treat cataract is accurately calculating the intraocular lens diopter. However, some patients did not feel satisfied because the error of calculation of intraocular lens diopter. With big data analysis and self-learning, artificial intelligence can deeply analyze and autonomously decide on complex clinical data. Therefore, this technology is expected to improve the calculation accuracy of intraocular lens diopter, to reduce postoperative refractive error and to improve patients' satisfaction. By referring to relevant literature at domestic and abroad, this paper is aimed to introduce the basic principle of artificial intelligence in intraocular lens diopter calculation, analyze and compare the characteristics, advantages and limitations of artificial intelligence based on different principles.