A Comparative Study of 2 Different Segmentation Methods of ADC Histogram for Differentiation Genetic Subtypes in Lower-Grade Diffuse Gliomas

Table 8

ROC curve analysis of different parameters for distinguishing IDH^mut/1p19q⁻ tumors and IDH^mut/1p19q⁺ tumors.

Parameter	ROI	AUC	Specificity (%)	Sensitivity (%)	Cutoff value	value
ADCmean	ROI-1	0.715	55.6	95.5	1546.32	0.151
	ROI-2	0.758	72.2	81.8	1387.97
P5 ADC	ROI-1	0.735	50.0	100.0	1108.00	1.00
	ROI-2	0.735	44.4	100.0	1110.00
P10 ADC	ROI-1	0.749	50.0	100.0	1177.50	1.00
	ROI-2	0.749	50.0	100.0	1191.00
P25ADC	ROI-1	0.766	66.7	86.4	1246.75	1.00
	ROI-2	0.766	50.0	100.0	1345.00
P50 ADC	ROI-1	0.763	55.6	100.0	1581.25	1.00
	ROI-2	0.763	67.7	86.4	1384.00
P75 ADC	ROI-1	0.710	50.0	100.0	1788.50	1.00
	ROI-2	0.710	66.7	81.8	1540.00
P90 ADC	ROI-1	0.669	44.4	90.9	1984.00	1.00
	ROI-2	0.669	55.6	90.9	1812.50
P95ADC	ROI-1	—	—	—	—	—
	ROI-2	0.689	61.1	86.4	1979.00
ADCmode	ROI-1	0.784	61.1	90.9	1448.75	0.729
	ROI-2	0.787	61.1	90.9	1450.00
Skewness	ROI-1	0.747	100.0	44.4	-0.164	0.178
	ROI-2	0.821	81.8	77.8	0.186
Kurtosis	ROI-1	—	—	—	—	—
	ROI-2	0.694	86.4	55.6	-0.349