In computed tomography (CT), artificial intelligence-based image reconstruction has become a revolutionary method for enhancing image quality and lowering radiation exposure.When compared to traditional filtered back projection and conventional iterative reconstruction, AI techniques, especially deep learning, enable superior noise reduction and artifact suppression by learning intricate mappings from noisy or undersampled projection data to high-quality images. For low-contrast lesion detection and overall diagnostic confidence, these models can maintain or even improve spatial resolution and noise texture. AI reconstruction has shown previously unheard-of performance in difficult situations like low-dose, sparse-view, and limited-angle CT, enabling clinically acceptable images from drastically reduced data.