Clinical and Biomarker-Based Predictors of Outcomes in Advanced Non-Small Cell Lung Cancer Patients Treated with First-Line Checkpoint Inhibitors With or Without Platinum-Based Chemotherapy

Obiettivo generale

• To generate adequate synthetic data from a real-world cohort of well-annotated consecutive LC patients using GANs and other generative models • To validate the synthetic data generated with a validation framework in terms of statistical fidelity, clinical utility and privacy preservability.

Risultati attesi

We expect to generate high-fidelity synthetic datasets that reliably reproduce the statistical distribution and clinical complexity of real-world NSCLC cohorts. Validated through the SAFE and MOSAIC frameworks, the models will demonstrate improved ability to predict outcomes in first-line immunother…