Application of Artificial Intelligence in Thoracic Diseases

In 2012, with the emergence of AlexNet in the ImageNet image classification competition, AI began to be increasingly applied in medical image research. A notable study was published in the journal Cell in 2018 [1]. In this study, deep learning models were first used to diagnose retinal diseases and differentiate pneumonia in children’s chest X-ray examinations using transfer learning. These models achieved good diagnostic performance, demonstrating that AI models can be applied to different types of diseases and can span both natural and medical image domains. Notably, this study was carried out by domestic scholars, indicating that China has occupied a place in the world’s AI medical research field. Since then, AI achievements in chest imaging have exponentially grown, reaching a new peak in 2023.