Skip to main content
. 2021 Sep 21;11:748229. doi: 10.3389/fonc.2021.748229

Table 3.

Preoperative UCSF eloquence compared to voxel-based eloquence in dLGG (N=277).

Presumed Eloquent (N=182) Non-eloquent (N=95) Measure of agreement P value1
Any voxel-based proximity of eloquent regions, No (%) 173 (95.1) 59 (62.1) 0.377 <0.001
No voxel-based proximity of eloquent regions, No (%) 9 (4.9) 36 (37.9)

1 Cohen’s Kappa test was used for interrater comparison between preoperative UCSF eloquence and voxel-based eloquence. Bold values indicate significant P value < 0.002.