Exploring radiomics and cytokine correlations in COVID-19: a study on lung damage, prognosis, and immune responses

This study utilizes artificial intelligence to refine the analysis of radiomic features from chest CT scans alongside cytokine profiles to predict clinical outcomes in COVID-19 patients. The study effectively correlates specific cytokines, such as IL-6, with lung damage severity, and these correlations are then used to predict hospitalization, mechanical ventilation, and mortality risks. The AI models developed provide a methodological advancement by linking radiomic data to predictive biomarkers, thereby supporting more accurate prognosis predictions in a clinical setting. This approach demonstrates the utility of integrating radiomic features and cytokine levels to assess the potential outcomes for COVID-19 patients, contributing to more targeted and informed clinical decision-making.