Information about heading direction is critical for navigation as it provides the means to orient ourselves in space. However, given that veridical head-direction signals require physical rotation of the head and most human neuroimaging experiments depend upon fixing the head in position, little is known about how the human brain is tuned to such heading signals. Here we adress this by asking 52 healthy participants undergoing simultaneous electroencephalography and motion tracking recordings (split into two experiments) and 10 patients undergoing simultaneous intracranial electroencephalography and motion tracking recordings to complete a series of orientation tasks in which they made physical head rotations to target positions. We then used a series of forward encoding models and linear mixed-effects models to isolate electrophysiological activity that was specifically tuned to heading direction. We identified a robust posterior central signature that predicts changes in veridical head orientation after regressing out confounds including sensory input and muscular activity. Both source localization and intracranial analysis implicated the medial temporal lobe as the origin of this effect. Subsequent analyses disentangled head-direction signatures from signals relating to head rotation and those reflecting location-specific effects. Lastly, when directly comparing head direction and eye-gaze-related tuning, we found that the brain maintains both codes while actively navigating, with stronger tuning to head direction in the medial temporal lobe. Together, these results reveal a taxonomy of population-level head-direction signals within the human brain that is reminiscent of those reported in the single units of rodents.

有关航向的信息对于导航至关重要，因为它提供了我们在空间中定位的方法。然而，考虑到真实的头部方向信号需要头部的物理旋转，并且大多数人类神经成像实验依赖于将头部固定在适当的位置，因此人们对人类大脑如何适应这种方向信号知之甚少。在这里，我们通过要求 52 名接受同步脑电图和运动跟踪记录的健康参与者（分为两个实验）和 10 名接受同步颅内脑电图和运动跟踪记录的患者来完成一系列定向任务，在这些任务中他们将头部物理旋转到目标位置来解决这个问题。然后，我们使用一系列前向编码模型和线性混合效应模型来隔离专门针对航向方向调整的电生理活动。我们确定了一个强大的后中央特征，可以预测在回归包括感觉输入和肌肉活动在内的混杂因素后真实头部方向的变化。源定位和颅内分析都表明内侧颞叶是这种效应的起源。随后分析从与头部旋转相关的信号和反映特定位置效应的信号中分离出头部方向特征。最后，当直接比较头部方向和与眼睛注视相关的调节时，我们发现大脑在主动导航时维持这两种代码，并且内侧颞叶对头部方向的调节更强。总之，这些结果揭示了人脑内群体水平头部方向信号的分类，这让人想起在单个啮齿类动物中报告的分类。