Following this puts new work involving it at the top of your briefing, with a note saying why it is there. Links are taken from the source record, never inferred.
Single-centre retrospective radiomics study with surrogate endpoint (pathologic complete response) and no prospective validation; promising methodology but limited generalizability and unclear sample adequacy.
A single-centre retrospective study with internal validation showing a diagnostic model for EGFR mutation prediction; the result is clear and clinically relevant but limited by retrospective design, modest sample size, and lack of external validation.
Single-centre prospective cohort with internal cross-validation only, no external validation, using a complex machine learning ensemble on a surrogate radiomic endpoint (week 3 imaging changes) rather than a direct clinical outcome; promising technical result but requires external validation before clinical use.