基于影像组学预测机器人辅助胃癌根治术患者预后的回顾性研究OA

Objective:To explore the predictive value of radiomics features for the prognosis of patients undergoing robot-assisted radical gastrectomy for gastric cancer.Methods:Clinical data from 60 patients who underwent Da Vinci robot-assisted radical gastrectomy at the Affiliated Hospital of Jiangnan University between January 2022 and December 2024 were retrospectively analyzed.Radiomics features of the lesions were extracted from arterial and venous phase CT images.Following feature selection,the least absolute shrinkage and selection operator(LASSO)-Cox regression model was used to construct a radiomics signature for predicting progression-free survival(PFS),and the Rad-score was calculated for each patient.Patients were stratified into high-risk and low-risk groups based on the median Rad-score.Univariate and multivariate Cox proportional hazards regression models were used to analyze independent prognostic factors influencing PFS.The accuracy and clinical utility of the model were validated using decision curve analysis(DCA).Results:Eight radiomics features significantly associated with PFS were selected and used to build the Rad-score.Kaplan-Meier analysis showed that the PFS in the high-risk group was significantly shorter than that in the low-risk group(P<0.05).Multivariate Cox analysis indicated that Rad-score(HR=3.452,95%CI:1.873-6.362,P<0.001)and TNM stage(Ⅲ vs.Ⅰ-Ⅱ,HR=2.987,95%CI:1.545-5.775,P=0.001)were independent predictors of PFS.The clinical-radiomics nomogram model constructed based on these factors showed the highest predictive performance,with the AUC for 1-year and 2-year PFS significantly higher than that of models using Rad-score or TNM stage alone.DCA showed that this nomogram model provided greater clinical net benefit for patient decision-making compared to"treat all"or"treat none"strategies.Conclusion:The preoperative contrast-enhanced CT-based radiomics signature is an effective tool for predicting the prognosis of gastric cancer patients after robot-assisted radical gastrectomy.The nomogram model combining radiomics and clinical factors can provide important reference for formulating individualized postoperative adjuvant treatment strategies.