Research Article

Using the K-Nearest Neighbor Algorithm for the Classification of Lymph Node Metastasis in Gastric Cancer

Table 2

Univariate analyses of the features of gastric lymph node metastasis arterial phase.

No.FeatureMean Standard 1 2 AUCSUIG
NegativePositive

140 keV 0.000 0.0000.5690.8750.1740.186
250 keV 0.000 0.0000.5400.8690.1740.186
360 keV 0.000 0.0020.4880.8450.1860.208
470 keV 0.000 0.0250.3620.7740.1060.104
580 keV 0.001 0.3020.1720.5960.0700.071
690 keV 0.025 0.994−0.0010.5020.0010.001
7100 keV 0.114 0.537−0.1030.5520.0130.015
8110 keV 0.272 0.295−0.1740.5990.0140.015
9120 keV 0.434 0.182−0.2210.6230.0250.027
10130 keV 0.570 0.127−0.2520.6530.0790.086
11140 keV 0.673 0.092−0.2770.6600.0790.086
12Effective-Z 0.000 0.0000.6010.8960.3170.336
13Calcium-Iodine 0.284 0.015−0.3910.7540.1260.127
14Calcium-Water 0.000 0.0000.5940.8990.3150.343
15Iodine-Calcium 0.000 0.0010.5000.8220.1740.165
16Iodine-Water 0.000 0.0000.5960.8960.3150.343
17Water-Calcium 0.000 0.002−0.4940.8180.1740.165
18Water-Iodine 0.291 0.017−0.3860.7340.1740.165