| High-resolution CT imaging features: solidation tumor ratio(CTR),disappearance-tumour-rate(DTR)and inflammation index: neutrophil-to-lymphocyte ratio(NLR)and systemic immune-inflammation index(SII)have been reported.lymphocyte ratio(NLR)and systemic immune-inflammation index(SII),which are readily available imaging and systemic immune inflammation response parameters,are associated with poor prognosis in lung cancer.However,studies on the prediction of the aggressiveness of these indices in ground-glass nodule(GGN)are limited.Objective: To analyze the aggressiveness of GGN by separate as well as combined diagnostic methods using CT imaging features and inflammatory metrics and to test the predictive efficacy.Methods: The data of patients with ground glass lung nodules with definite pathological findings who had completed surgery or puncture biopsy in Taizhou People’s Hospital from September 2018 to May 2022 were retrospectively analyzed.Based on SPSS 26.0,Graph Pad Prism 9.0,and R Studio software.The age,gender,blood routine,tumor markers and CT imaging data of the two groups were compared,and the indicators that were statistically significant in the single-factor analysis were included in the multi-factor logistic regression analysis by logistic single-factor regression analysis,after excluding the covariance,we first used the independent influencing factors in the multi-factor analysis individually and jointly for the subject working characteristics Receiver operating characteristic curve(ROC)curve analysis was performed to comprehensively assess the difference in the efficacy of single and combined tests in diagnosing GGN as invasive or non-invasive.The multifactorial analysis of significant indicators was then imported into R software,and the column line graph model was plotted using the rms data package.1000 replicate samples were taken using the Bootstrap method to internally validate the prediction model,and the model was evaluated by discrimination and calibration.The p ROC data package was used to plot the ROC curve and calculate the C-index parameter to assess the discrimination of the model,and the calibration was evaluated by using the Calibration Plot.Results: This study showed that SII,lung window tumor maximum diameter(D-MAX)and DTR were independent influences on the invasive occurrence of GGN,and the area under cueve(AUC)for the diagnosis of lung carcinogenesis was 0.728(95% CI: 0.666-0.791)for SII,D-MAX and DTR when tested alone,respectively.0.779(95% CI: 0.720-0.839)and 0.706(95% CI: 0.641-0.771),all with P<0.05;the sensitivity was 76.4%,76.4% and 85.4%,and the specificity was 58.1%,65.5% and49.4%,respectively.The area under the curve AUCs of SII+D-MAX,SII+DTR,D-MAX+DTR and SII+D-MAX+DTR combined for the diagnosis of lung carcinogenesis were 0.820(95% CI: 0.766-0.874),0.779(95% CI: 0.721-0.836)and0.782(95% CI: 0.723-0.840),0.825(95% CI: 0.773-0.878),all p < 0.05;sensitivity was 64.0%,80.9%,60.7% and 83.2%,and specificity 86.2%,62.6%,84.5% and68.4%,respectively.The column line graph model,which was internally validated using the Boostrap method,had a C-index of 0.793(95% CI: 0.7366-0.8489),and the calibration curve showed good performance of the model.Conclusion: The area under the curve AUC of the combined SII+D-MAX+DTR indicators for the diagnosis of GGN invasive onset was higher than that of both the individual test and the two combinations,with higher diagnostic efficacy.Specifically,among the seven indicators,although the sensitivity of the single indicator DTR was the largest,it was only 2.2% higher than that of SII+D-MAX+DTR.However,the specificity,positive predictive value,and negative predictive value of DTR were much smaller than that of SII+D-MAX+DTR,and collectively,the combined prediction,i.e.,SII+D-MAX+DTR,was more advantageous in predicting lung carcinogenesis.The established column line graph model has high discrimination and accuracy,which is a guideline for screening the GGN population and formulating corresponding intervention strategies. |
PDF Full Text Request