Abstract:

Objective: To investigate the electrocorticographic signal activity characteristics in glioma located in the motor area. Methods: Total 8 patients undergoing awake craniotomy for motor area- involved glioma in West China Hospital, Sichuan University between October 2024 and February 2025 were enrolled. During the awake state, 2 min resting-state electrocorticographic signals were continuously recorded. After preprocessing steps including filtering, signal activity characteristics across θ, α, β, γ1, γ2, γ3, γ4 frequency bands were extracted. Using preoperative MRI data and subdural electrode parameters, individualized brain and electrode models were reconstructed. The spatial relationships between electrodes and tumors during data acquisition were coregistered onto the brain models. Based on the tumor invasion degree into brain tissue beneath the electrodes, the 762 validated electrodes were categorized into 3 groups: normal brain tissue group (n = 460), tumor-invaded cortex group (n = 274) and tumor-non-invaded cortex group (n = 28). Results: In the frequency bands of θ (H = 249.993, P = 0.000), α (H = 251.311, P = 0.000), β (H = 288.834, P = 0.000), γ1 (H = 312.145, P = 0.000), γ2 (H = 263.777, P = 0.000), γ3 (H = 238.691, P = 0.000), γ4 (H = 208.830, P = 0.000), the differences in the electrocorticographic signal activity characteristics among the 3 groups were statistically significant. Further pairwise comparisons revealed that the electrocorticographic signal activity characteristics of the normal brain tissue group were higher than those of the tumor-invaded cortex group in the frequency bands of θ (Z = 5.711, P = 0.000), α (Z = 5.823, P = 0.000), β (Z = 6.907, P = 0.000), γ1 (Z = 7.286, P = 0.000), γ2 (Z = 6.054, P = 0.000), γ3 (Z = 5.247, P = 0.000), γ4 (Z = 4.647, P = 0.000), and higher than those of the tumor-non-invaded cortex group in the frequency bands of θ (Z = 4.051, P = 0.000), α (Z = 3.277, P = 0.001), β (Z = 4.172, P = 0.000), γ1 (Z = 5.013, P = 0.000), γ2 (Z = 4.749, P = 0.000), γ3 (Z = 4.264, P = 0.000). The electrocorticographic signal activity characteristics of the tumor-non-invaded cortex group were higher than those of the tumor-invaded cortex group in the frequency bands of θ (Z =-2.071, P = 0.038), α (Z =-2.871, P = 0.004), β (Z =-2.403, P = 0.016). Conclusions: This study found that normal brain tissue exhibits higher activity characteristics in electrocorticographic signals compared to glioma tissue. These findings suggest the potential for developing glioma localization techniques based on electrocorticographic signals, which could assist neurosurgeons in achieving more precise glioma resection.

Key words: Glioma, Motor cortex, Electrocorticography

摘要：

目的: 探讨位于运动区的胶质瘤皮质脑电信号活动度特征。方法: 纳入2024年10月至2025年2月在四川大学华西医院接受唤醒手术且胶质瘤累及运动区的8例患者，于术中清醒状态下连续采集2 min的静息态皮质脑电信号，提取θ、α、β、γ1、γ2、γ3、γ4频段的脑电信号活动度特征。基于术前MRI数据和电极参数重建脑模型与电极矩阵，并将数据采集过程中电极与肿瘤的位置关系配准至脑模型，根据电极下肿瘤对脑组织的侵袭程度将最终纳入的762个有效电极分为3组，即正常脑组织组（460个）、肿瘤侵入皮质组（274个）和肿瘤未侵入皮质组（28个）。结果: 3组皮质脑电信号活动度特征在θ（H=249.993，P=0.000）、α（H=251.311，P=0.000）、β（H=288.834，P=0.000）、γ1（H=312.145，P=0.000）、γ2（H=263.777，P=0.000）、γ3（H=238.691，P=0.000）、γ4（H=208.830，P=0.000）频段范围内差异均具有统计学意义。进一步两两比较，正常脑组织组皮质脑电信号活动度特征在θ（Z=5.711，P=0.000）、α（Z=5.823，P=0.000）、β（Z=6.907，P=0.000）、γ1（Z=7.286，P=0.000）、γ2（Z=6.054，P=0.000）、γ3（Z=5.247，P=0.000）、γ4（Z=4.647，P=0.000）频段均高于肿瘤侵入皮质组，在θ（Z=4.051，P=0.000）、α（Z=3.277，P=0.001）、β（Z=4.172，P=0.000）、γ1（Z=5.013，P=0.000）、γ2（Z=4.749，P=0.000）、γ3（Z=4.264，P=0.000）频段均高于肿瘤未侵入皮质组；肿瘤未侵入皮质组在θ（Z=-2.071，P=0.038）、α（Z=-2.871，P=0.004）、β（Z=-2.403，P=0.016）频段均高于肿瘤侵入皮质组。结论: 正常脑组织的皮质脑电信号活动度特征高于胶质瘤组织，提示未来可以开发基于皮质脑电信号的胶质瘤位置识别技术，以辅助神经外科医师进行胶质瘤的精准切除。

关键词: 神经胶质瘤, 运动皮质, 脑皮层电图