关键词: 肺肿瘤, 弥散磁共振成像, 曲线下面积

Abstract: ［Abstract］ Objective〖HTSS〗〓To explore the value of the wholelesion apparent diffusion coefficient(ADC) value histogram of DWI examination in the differentiation of malignant and benign lung lesions, and choose the best diagnostic parameters.
〖HTH〗〖WTHZ〗Methods〖HTSS〗〓The 22 lung cancer and 17 benign lung lesions (including pneumonia and tuberculosis), which received lung biopsy or surgical operation and had pathology results were collected. Before treatment, all patients received lung DWI examination with 6 b values: b=0, 50, 150, 400, 600, 800 s/mm2. The wholelesion ADC value histogram and related parameters was calculated, including skewness, kurtosis, entropy, ADC mean value, ADC median value, lesion volume, ADC value standard deviation, and the 5 th, 10 th, 15 th, 30 th, 50 th, 70 th ,90 th ADC value percentile. And the differences between the 2 groups were compared.
〖HTH〗〖WTHZ〗Results〖HTSS〗〓Lung cancer group parameters of ADC values histogram of skewness, kurtosis, mean, standard deviation, ADC50th, ADC70th, ADC90th were lower than lung benign lesions, the difference was statistically significant(P＜005).  Receiver operator characteristic curve(ROC) was plotted with ADC histogram parameters to calculate area under curve(AUC). The diagnostic efficacy of the ADC standard deviation was the highest. When the ADC standard deviation was 0.810(×10-3mm2 /s), the AUC was 0.789, the specificity was 94.1%, and the sensitivity was 68.2%. When ADC90th is 1.940(×10-3mm2/s), AUC is 0.785, specificity is 94.1%, and sensitivity is 54.5%.
〖HTH〗〖WTHZ〗Conclusion〖HTSS〗〓The wholelesion apparent diffusion coefficient(ADC) value histogram can be used in the differentiation of malignant and benign lung lesions. Among the parameters, ADC value standard deviation have the best diagnostic performance.

Key words: lung neoplasms, diffusion weighted imaging, area under curve