Skip to main content
. 2020 Feb 25;10:198. doi: 10.3389/fonc.2020.00198

Table 3.

The diagnostic performance of texture features derived from simutanous18F-FDG PET image and the ADC map for predicting synchronous distant metastasis in pancreatic ductal adenocarcinoma patients.

Texture feature M0 group M1 group P1 AUC 95% confidence intervals P2 Optimal cut-off value
PET-GLRLM_RLNU 467 ± 173 1391 ± 738 0.002 0.939 0.711–0.998 <0.0001 >751
PET-GLRLM_LRHGE 172 ± 77 288 ± 96 0.037 0.818 0.560–0.960 0.005 >120.6
PET-NGLDM_Coarseness 0.015 ± 0.008 0.007 ± 0.003 0.037 0.803 0.543–0.952 0.034 < =0.01
PET-GLZLM_GLNU 5.4 ± 2.3 11.3 ± 6.8 0.020 0.848 0.595–0.973 0.0004 >8.3
ADC-HISTO_Skewness −0.02 ± 0.59 0.72 ± 0.80 0.048 0.795 0.534–0.948 0.001 >0.14
ADC-HISTO_Kurtosis 3.32 ± 1.16 5.25 ± 2.01 0.048 0.803 0.543–0.952 0.008 >3.84
ADC-GLRLM_LRE 36 ± 10 66 ± 24 0.007 0.894 0.650–0.989 <0.0001 >42.98
ADC-GLRLM_SRHGE 786 ± 51 654 ± 131 0.048 0.803 0.543–0.952 0.006 < =691
ADC-GLRLM_LRLGE 0.008 ± 0.002 0.019 ± 0.009 0.005 0.879 0.631–0.984 <0.0001 >0.012
ADC-GLRLM_LRHGE 0.15E+6 ± 0.04E+6 0.28E+6 ± 0.10E+6 0.007 0.894 0.650–0.989 <0.0001 >0.18E+6
ADC-GLRLM_GLNU 227 ± 96 442 ± 142 0.003 0.924 0.690–0.996 <0.0001 >269.4
ADC-GLRLM_RLNU 30 ± 8 45 ± 9 0.005 0.909 0.670–0.993 <0.0001 >32.6
ADC-GLRLM_RP 0.212 ± 0.023 0.165 ± 0.033 0.007 0.879 0.631–0.984 <0.0001 <0.18
ADC-GLZLM_LZE 1.53E+6 ± 1.67E+6 9.50E+6 ± 8.71E+6 0.010 0.879 0.631–0.984 <0.0001 >4.62E+6
ADC-GLZLM_LZLGE 362 ± 395 2249 ± 2061 0.010 0.879 0.631–0.984 <0.0001 >1094.1
ADC-GLZLM_LZHGE 0.65E+10 ± 0.70E+10 4.01E+10 ± 3.68E+10 0.010 0.879 0.631–0.984 <0.0001 >1.95E+10
ADC-GLZLM_ZP 0.0016 ± 0.0017 0.0006 ± 0.0004 0.050 0.795 0.534–0.948 0.010 < =0

ADC, apparent diffusion coefficient; AUC, area under receiver operating characteristic (ROC) curve; FDG, fluorodeoxyglucose; M0, no synchronous distant metastasis; M1, with synchronous distant metastasis; NA, not applicable; P1, P-value for independent- samples Mann-Whitney U-test, indicates p-value is significant <0.05; P2, p-value for AUC, indicates P-value is significant <0.05; PET, positron emission tomography; 95% CI, 95% confidence interval; GLRLM, Gray level run length matrix; GLZLM, Gray level zone length matrix; SRHGE, Short-nun high gray-level emphasis; GLNU, Gray-level non-uniformity; RLNU, Run length non-uniformity; LZE, Long-zone emphasis; NGLDM, Neighborhood gray-level different matrix; SZHGE, Short-zone high gray-level emphasis; LZHGE, Long-zone high gray-level emphasis.