Construction and validation of a CT-based radiomics-deep learning signature for non-invasive prediction of PD-L1 expression and immunotherapy outcomes in non-small cell lung cancer

Key Points

• The predictive ability of RADLsig shows promise for estimating PD-L1 expression status in patients.
• The analysis indicates RADLsig could aid in selecting patients for immunotherapy with improved outcomes.
• Utilization of deep learning techniques on CT images was central to the predictive model's development.
• Highlighting a non-invasive approach, this method may enhance personalized treatment strategies in clinical settings.

Abstract