Abstract
OBJECTIVE: Our study was aimed to established imaging models that can predict the prognosis of breast cancer patients based on IL18, so as to facilitate the formulation of more individualized treatment plans during the treatment. METHOD: We established enhancing MRI models to predict the mRNA expression of IL18 in breast cancer tissues. Meanwhile, bioinformatics analysis was integrated to studied the expression of IL18 in breast cancer with its clinical significance to explore the potential molecular mechanisms and its association with the immune microenvironment. RESULT: In breast cancer patients, high levels of IL18 expression are associated with improved overall survival (OS), and natural killer cells (NK cells) significantly increased (P < 0.001) with elevated IL18. Three radiomics models, RFE-GBM, Relief-GBM, and RFE_Relief-GBM were established in our study, each model demonstrated strong predictive consistency in the prognosis of breast cancer patients. CONCLUSION: IL18 is a protective factor in breast cancer, our radiomics models can serve as valuable non-invasive tools for utilizing mRNA expression of IL18 by enhancing MRI and show promise in predicting prognosis of breast cancer patients. CLINICAL TRIAL NUMBER: Not applicable.