Diagnostic models of solitary pulmonary mass lesion based on PET metabolic parameters
10.13929/j.1003-3289.201812012
- Author:
Wei JIANG
1
Author Information
1. Department of Molecular Imaging and Nuclear Medicine, Tianjin Medical University Cancer Institute and Hospital, National Clinical Research Center for Cancer, Key Laboratory of Cancer Prevention and Therapy, Tianjin's Clinical Research Center for Cancer
- Publication Type:Journal Article
- Keywords:
Fludeoxyglucose F 18;
Lung neoplasms;
Positron-emission tomography;
Support vector machine
- From:
Chinese Journal of Medical Imaging Technology
2019;35(5):696-700
- CountryChina
- Language:Chinese
-
Abstract:
Objective: To establish mathematical prediction models based on PET metabolic parameters, and to explore their value for differentiating benign and malignant solitary pulmonary lesions. Methods Data of 135 patients with solitary pulmonary lesions who underwent 18F-FDG PET/CT scan were retrospectively analyzed. PET metabolic parameters of the lesions were obtained, including metabolic tumor volume (MTV), maximum standardized uptake value (SUVmax), peak standardized uptake value (SUVpeak), mean standardized uptake value (SUVmean) and total lesion glycolysis of standardized uptake value (SUVtlg), as well as parameters of standardized uptake normalized to lean body mass (SUL), including SULmax, SULpeak, SULmean and SULtlg. The parameters above were used to establish support vector machine (SVM) models, which were selected according to the Akaike's information criterion (AIC). The diagnostic performances of the models were assessed with ROC curves. The permutation test was used for internal validation. Results: Two sets of optimization models were obtained and recorded as Mgroup A (include MTV, SUVpeak and SUVtlg) and Mgroup B (include MTV, SUVpeak and SULtlg). AUC of Mgroup A model was 0.865 (P=0.021), with the sensitivity of 82.72%, specificity of 83.33% and diagnostic accuracy of 82.96%, of Mgroup B model was 0.863 (P=0.030), with the sensitivity of 82.72%, specificity of 83.33% and diagnostic accuracy of 82.96%, respectively. There was no statistically significant difference of AUC between the two models (P=0.294). Both models were reliable evaluated with the permutation test. Conclusion: SVM models based on PET metabolic parameters can be used for differential diagnosis of benign and malignant solitary pulmonary lesions, whereas metabolic parameters corrected by lean body mass bring no remarkable improvement on diagnostic efficacy.
