Objective:To construct and verify a prediction and evaluation model based on dual-energy computed tomography(CT)radiomics for lymph node metastasis in gastric cancer,so as to provide more accurate and reliable method before lymph node in gastric cancer occurs metastasis.Methods:A total of eighty patients with gastric cancer admitted to Handan Central Hospital from January 2021 to December 2024 were retrospectively selected,and they were divided into a modeling group(48 cases)and a verification group(32 cases)using a random number table method according to a ratio of 3 to 2.The influencing factors were analyzed,and predictive model was constructed and verified by using univariate and binary logistics regression.The patients'general information,tumor size,differentiation degree,lesion location,and carbohydrate antigen 125(CA125),carbohydrate antigen 199(CA199)and carcinoembryonic antigen(CEA)were compared between two groups.Results:There were not significant differences in age,gender,body mass index(BMI),tumor size,degree of differentiation,lesion location in the modeling group(P>0.05).In the modeling group,the immunity concentration(IC)was(20.66±2.85)μg/mL,and normal immunity concentration(nIC)was(0.45±0.06)μg/mL,and the slope was(3.52±0.30).The three indicators were respectively(21.09±3.25)μg/mL,(0.47±0.05)μg/mL,(3.49±0.42)in verification group.There were significant differences in the three indicators among two groups(t=3.277,5.287,2.918,P<0.05).There were not significant differences in the tumor-related detection indicators between two groups(P>0.05).The results of binary logistics regression analysis showed that IC,nIC,and slope were influential factors in lymph node metastasis of gastric cancer(OR=2.564,1.647,1.786,P<0.05).The constructed prediction model by using Python Scikit-learn showed favorable appearance in the calibration curve with a slope closed to 1,which indicated a high degree of consistency between predicted risks and actual risks.Receiver operating characteristic(ROC)curve analysis showed that the area under curve(AUC)of the modeling group was 0.896(95%CI:0.8154～0.9167),and the sensitivity and specificity were respectively 84.60%and 80.69%.The above three indicators were respectively 0.853(95%CI:0.7982～0.8671),86.66%and 80.00%in the verification group.Conclusion:A prediction and evaluation model for lymph node metastasis of gastric cancer based on dual-energy CT radiomics is successfully constructed,and its predictive efficiency and clinical application value are verified.

Objective:To construct and verify a prediction and evaluation model based on dual-energy computed tomography(CT)radiomics for lymph node metastasis in gastric cancer,so as to provide more accurate and reliable method before lymph node in gastric cancer occurs metastasis.Methods:A total of eighty patients with gastric cancer admitted to Handan Central Hospital from January 2021 to December 2024 were retrospectively selected,and they were divided into a modeling group(48 cases)and a verification group(32 cases)using a random number table method according to a ratio of 3 to 2.The influencing factors were analyzed,and predictive model was constructed and verified by using univariate and binary logistics regression.The patients'general information,tumor size,differentiation degree,lesion location,and carbohydrate antigen 125(CA125),carbohydrate antigen 199(CA199)and carcinoembryonic antigen(CEA)were compared between two groups.Results:There were not significant differences in age,gender,body mass index(BMI),tumor size,degree of differentiation,lesion location in the modeling group(P>0.05).In the modeling group,the immunity concentration(IC)was(20.66±2.85)μg/mL,and normal immunity concentration(nIC)was(0.45±0.06)μg/mL,and the slope was(3.52±0.30).The three indicators were respectively(21.09±3.25)μg/mL,(0.47±0.05)μg/mL,(3.49±0.42)in verification group.There were significant differences in the three indicators among two groups(t=3.277,5.287,2.918,P<0.05).There were not significant differences in the tumor-related detection indicators between two groups(P>0.05).The results of binary logistics regression analysis showed that IC,nIC,and slope were influential factors in lymph node metastasis of gastric cancer(OR=2.564,1.647,1.786,P<0.05).The constructed prediction model by using Python Scikit-learn showed favorable appearance in the calibration curve with a slope closed to 1,which indicated a high degree of consistency between predicted risks and actual risks.Receiver operating characteristic(ROC)curve analysis showed that the area under curve(AUC)of the modeling group was 0.896(95%CI:0.8154～0.9167),and the sensitivity and specificity were respectively 84.60%and 80.69%.The above three indicators were respectively 0.853(95%CI:0.7982～0.8671),86.66%and 80.00%in the verification group.Conclusion:A prediction and evaluation model for lymph node metastasis of gastric cancer based on dual-energy CT radiomics is successfully constructed,and its predictive efficiency and clinical application value are verified.