Application value of spiral CT combined with autoantibody tests in the diagnosis of early lung cancer

This study is for exploring the diagnostic value of spiral CT combined with autoantibody tests for early lung cancer. In this paper, the total 340 samples is selected, which is divided into 3 groups: 140 early lung cancer patients, 100 patients with benign lung diseases and 100 healthy individuals for physical examination. Three groups of subjects were subjected to multi-slice spiral CT examination and 7 items of autoantibody examination, and the multi-slice spiral CT signs and 7 items of autoantibody levels were observed and compared. Based on the results of the two examinations, a binary logistic regression model was constructed, and this study conducted ROC curve analysis on its value in the diagnosis of early lung cancer. Out of 340 study subjects included in this study, 18 cases had missing clinical data >5%, and 322 valid data were ultimately collected, with a valid data recovery rate of 94.71%. The statistically significant difference did not exist in the halo sign, cavity sign, and Broncho vascular bundle sign among the three groups of subjects on multi-slice spiral CT (P > 0.05), while the statistically significant difference existed in the vacuolar sign, calcification sign, bronchial inflation signs, hairpin sign, lobulation sign, pleural indentation sign, and ground glass nodule sign (P < 0.05). The diagnostic results of multi-slice spiral CT showed 97 cases of benign lung lesions and 130 cases of lung cancer. The statistically significant differences existed in the levels of p53, PGP9.5, SOX2, GAGE7, GBU4-5, MAGE A1, and CAGE among the three groups of subjects (P < 0.05). Based on the results of multi-slice spiral CT and 7 autoantibody tests, binary logistic regression models F1 and F2 were constructed. The Omnibus test results showed that model χ2 = 254.004, P < 0.001. −2 log likelihood = 52.490, Cox &amp; Snell R2 = 0.673, Nagelkerke R2 = 0.909, Hosmer Limeshow test χ2 = 3.311, P = 0.913, indicating good model fit. The ROC curve analysis results showed that multi-slice spiral CT with 7 individual autoantibodies, 7 combined autoantibodies, and 7 combined autoantibodies had good predictive value for early lung cancer (AUC = 0.645–0.989, P < 0.05), with model F2 having the highest predictive value (AUC = 0.989, P < 0.05). Multi-slice spiral CT and 7 autoantibody tests have high diagnostic value for early lung cancer, and combined examination can further improve diagnostic efficiency.