Research progress and prospects of AI+ empowering chest X-ray and CT in the diagnosis and treatment of lung diseases

Lung diseases have long been at the forefront of global mortality and disability, although chest X-ray and CT are the basic entrances for screening, diagnosis and follow-up, they are exposed to limitations such as miss diagnosis, misdiagnosis and insufficient quantification under high load and complex disease spectrum. The rise of deep learning, radiomics, and multimodal large models has made Artificial Intelligence (AI) a key driving force for chest images to move from "reading tools" to "system engineering". AI has significantly improved detection, segmentation, phenotypic quantification, and risk prediction capabilities in multi-spectrum tasks such as lung nodules/lung cancer, tuberculosis, pneumonia, interstitial lung disease (ILD), chronic obstructive pulmonary disease (COPD), small airways, and pulmonary vascular diseases, and has stabilized key indicators such as doubling time, fibrosis burden, and airway remodeling, becoming an important technical basis for the implementation of Fleischner, American College of Chest Physicians (ACCP), and China guidelines. In prevention and screening, AI supports the identification of high-risk groups, large-scale chest X-ray screening, LDCT risk stratification, and early detection of subclinical abnormalities such as ILA and small airway disease, which can be combined with health management, digital twins, and metaverse platforms to build a forward-moving defense line intervention model. Physicians and patients generate structured reports, provide "guide online" decision support, and output differentiated explanations by using imaging diagnostic models and medical GPTs. AI also empowers radiotherapy planning, preoperative navigation, treatment response prediction, and lung function estimation, promoting image-function integration and individualized long-term management for treatment and follow-up. In the future, it will focus on the construction of general chest imaging large models, the deep integration of 5P medicine, the construction of federated learning and global collaborative data networks, and move from "intelligent imaging links" to the whole course of the disease system project that connects "hospital-community-family-cloud-metaverse", so that chest X-ray and CT will become the key infrastructure of the digital respiratory health ecosystem.

Key Words: AI; Lung cancer screening; radiomics and multimodal foundation models; quantitative phenotyping of ILD and COPD; digital twin and metaverse medicine; medical GPT and intelligent decision support