Objective To develop a predictive model for the lymph node ratio (LNR) in gastric cancer patients based on radiomics features extracted from preoperative imaging, and to construct a nomogram incorporating clinical factors to preoperatively predict LNR status. This model aims to assist in making more precise clinical decisions for gastric cancer treatment. Methods Clinical data of patients(n=380) who underwent radical gastrectomy at the First Affiliated Hospital of Nanjing Medical University between January 2014 and December 2018 were retrospectively collected. Patients with pathologically confirmed lymph node metastasis were selected. Preoperative imaging data were analyzed to extract radiomic features, followed by machine learning modeling. Three algorithms—Support Vector Machine (SVM), Random Forest (RF), and Logistic Regression (LR)—were evaluated. The RF model, which demonstrated the best performance based on the ROCcurve analysis, was selected for prediction. A nomogram was then developed using the RF score and clinically significant variables identified by multivariate logistic regression analysis. Results LNR was identified as an independent prognostic factor for patients undergoing gastric cancer resection. Three variables—RF score (p < 0.01), sex (p = 0.025), and clinical N stage (p < 0.01)—were included in the final nomogram model. The model’s performance was evaluated using ROC (The AUC value = 0.818)and decision curve analysis (DCA), both of which demonstrated good predictive ability. Conclusion The nomogram model based on radiomic and clinical information provides favorable predictive accuracy for preoperative LNR status in gastric cancer patients.