AI-guided Prognostication and Cranial Radiotherapy Optimization in EGFR-TKI-treated Non-small Cell Lung Cancer Patients With Baseline Brain Metastases

The goal of this observational study is to extract the imaging features of brain lesions and primary lung lesions in NSCLC patients with brain metastases by deep learning, as well as common clinicopathological parameters, which are used to construct a multimode model that can accurately predict the treatment efficacy and survival of the third-generation EGFR-TKI treatment, and to use the model to assist in screening high-risk populations suitable for upfront cranial radiotherapy. Participants receiving third-generation EGFR-TKI treatment will be enrolled in our study and we will collect their regular contrast-enhanced chest CT and contrast-enhanced brain MRI for model construction.