MRI, DTI, EEG, MEG, fMRI, MVPA

Structural magnetic resonance imaging (MRI)

• MRI is detecting content of hydrogen atoms in tissue
  • It will detect the radio frequency signal of resonance frequency

Diffusion Tensor Imaging (DTI)

• DTI measures the direction and magnitude of water diffusion in the brain
  • in isotropic and anisotropic environments

ElectroEncephaloGraphy

• Non-invasive, Can detect the spontaneous electrical activity of the brain
  • Which represent the postsynaptic potentials of pyramidal neurons in the neocortex and allocortex
  • Can not detect activity in deep brain structures

MagnetoEncephaloGraphy

• Signal derives from net effect of ionic currents during synaptic transmission

Comparison: EEG, MEG, fMRI

• Cost: EEG < MEG/ fMRI
• Spatial resolution: EEG < MEG < fMRI
• Temporal resolution: fMRI < EEG/ MEG
  • 说明EEG和MEG能够更快地捕捉神经活动的变化，而fMRI的时间分辨率低

Functional Magnetic Resonance imaging

• BOLD (blood oxygenation level dependent effect)
  • 通过血氧情况，分析代谢，活动等情况
  • 并不能直接和neural activity相关，只能correlated to it
• fMRI的关键概念：
  • Voxels：三维空间的体积像素
  • planes: like coronal plane, horizontal plane, Sagittal plane
  • 通过参考标准化的坐标系统，确定不同区域的位置：
    • Montreal Neurological Institute space (152 normal scans, average)
    • Talairach space: one dissected and photographed brain ![[截屏2025-03-31 10.32.33.png]]
• 统计推断
  • Betas: can be calculated from raw data, which represents the activity strength of specific regions
    • calculated by generalized linear model
    • Beta值代表了大脑区域对特定条件的激活强度
    • 这些检验帮助确定了大脑区域在特定任务中显著活跃
  • Two types of corrections:
    • Whole brain correction
    • Small volume correction
• Two analysis method in fMRI
  • Mass-univariate approach
    • conventional
    • Data is smoothed and averaged across subjects
    • Can analyze groups differences voxel by voxel, based on average activity level
  • Multi-voxel pattern analysis (MVPA)
    • 行为预测基于局部或者全局
    • 两种不同的刺激可能激活相同的脑区(平均活动相似)，但激活模式不同，MVPA能够区分这些模式并将其与不同的认知状态或刺激关联起来。