Objective To predict the radiation resistance in patients with triple-negative breast cancer based on spectral CT imaging parameters and combined with vitamin K deficiency or antagonist Ⅱ-induced protein (PIVKA-Ⅱ）.Methods A total of 80 patients with triple-negative breast cancer who received radiotherapy in our hospital from May 2021 to May 2024 were selected and their general data and energy spectrum CT imaging characteristics were analyzed. Logistic regression was used to analyze the factors affecting the patients' radiotherapy resistance. Logistic regression was used to analyze the relationship between energy spectrum CT imaging parameters and PIVKA-Ⅱ. Logistic regression algorithm was used to establish the radiation resistance prediction model of energy spectrum CT imaging parameters combined with PIVKA-Ⅱ, ROC curve and AUC to evaluate the prediction efficiency. Results There were no significant differences between the two groups in cancer location, body mass index (BMI), menstrual status, hypertension, diabetes, smoking history, drinking history, age at first delivery, family history, BRCA mutation, lymph node metastasis, type of surgery, CEA, CA153, etc. (P>0.05). There were statistical differences in age, pathological grade, course of disease, long-term use of contraceptives or estrogen, PIVKA-Ⅱ (P<0.05). Among the characteristics of energy spectrum CT imaging parameters, there were statistical differences in k value (40~70keV) in arterial and venous phases, IC value in arterial and venous phases, focal diameter and tumor margin between the two groups (P<0.05). Age, pathological grade, course of disease, long-term use of contraceptives or estrogen, PIVKA-Ⅱ concentration were the risk factors for radiotherapy resistance (OR>1, P<0.05), and there was no multicollinearity among the variables. After adjusting for age, pathological grade, course of disease, long-term use of contraceptives or estrogen factors, there was a correlation between arterial k40 70keV, arterial IC, venous k 40-70kev,venous IC and focal length and PIVKA-Ⅱ concentration (P<0.05). All parameters of energy spectrum CT imaging combined with PIVKA-Ⅱ had good predictive value, among which PIVKA-Ⅱ+arterial phase k 40~70keV +venous phase k 40~70keV +arterial phase IC+venous phase IC+lesion length diameter (P4) had the highest predictive efficiency, and the AUC was 0.855. Conclusion Energy spectrum CT image parameters combined with PIVKA-Ⅱ has a good value in