Abstract: Navigation be classified into two different types, piloting and path integration. Path integration refers to the process that navigators integrate information regarding self-motion to estimate their current position and orientation relative to the origin of their travel. Path integration is common to many species, such as insects, birds, and mammals. In particular, humans show path integration abilities in path completion tasks in which they travel along several segments and then attempt to directly return to the origin. In order to perform path integration, humans may rely on pure internal information, such as body senses from the vestibular, proprioceptive, and efferent systems。 Humans can also reply on pure external information, such as optic flow, or a mixture of both internal and external cues, to perform path integration. Many behavioral studies of human path integration have been focusing on the spatial representation and spatial updating. A pure egocentric representation is regarding the position of surroundings relative to the navigator, which means that the representation needs to be updated whenever the navigator moves. If human path integration is supported by egocentric representations, navigators need to continuously update the original relative to their own position and orientation without having a detailed representation of the outbound paths. In contrast, a pure allocentric representation is independent of the navigator and defined relative to the surroundings or the earth. Thus, when the navigator moves, the representation does not change. If human path integration is supported by allocentric representations, the navigators need to have a detailed representation of the outbound paths. The literature has shown that both egocentric and allocentric representations can support human path integration, but future research is needed to test whether humans can switch between these two types of spatial updating strategies. Recent neuroscience studies have shown that the medial temporal lobe structures and the parietal lobe structures might be relevant to human path integration. Although animal studies showed that the hippocampus and medial entorhinal cortex are essential for path completion performance, human lesion studies provided mixed evidence regarding the role of the medial temporal lobe in human path integration. That is, some studies have revealed that the medial temporal lobe is important to human path integration, but others suggested that human path integration might not rely on the medial temporal lobe. On the other hand, some studies have revealed that the interaction between human motion complex and regions is responsible for spatial cognition, such as hippocampus and medial prefrontal lobe, plays a role in human path integration based on optic flow. Furthermore, a neural model of spatial memory suggested that the parietal lobe is responsible for maintaining egocentric representations and therefore might be important to human path integration. Empirical evidence has also shown high correlations between the egocentric spatial updating and the activation in the precuneus and dorsal premotor cortex. Previous studies have revealed that humans show worse performance in the path completion task than nonhuman animals. Future research is needed to examine whether humans have limited path integration abilities or whether it is due to methodological difference in human and nonhuman path integration studies. On the other hand, previous studies have provided some preliminary evidence regarding the influence of non-sensory factors on human path integration, but more research is needed to further examine these effects, in particular, the effects of expectation, knowledge, and experience on human path integration. Recently, Virtual Reality is a powerful tool for human path integration studies because of two reasons. For one, it allows researchers to control the navigation environment by manipulating landmarks and surface texture of the roads. For another, it also allows researchers to conduct fMRI or ERP studies to further investigate the neural mechanisms of human path integration.

Key words: path integration, navigation, spatial representation, spatial updating

摘要： 路径整合是指巡航者对与自身运动有关的信息进行整合来完成巡航任务的过程。这些与自身运动有关的信息可以是内源性的，如前庭感觉、本体感觉、动作指令的信息；也可以是外源性的，如视觉流。路径整合在许多物种中存在。人类路径整合的行为实验表明，以自我为参照系的空间表征和以环境为参照系的空间表征都有可能支持路径整合。神经科学的研究则表明，海马、内嗅皮层等内侧颞叶区域和以楔前叶等顶叶区域都与人类路径整合密切相关。

关键词: 路径整合, 巡航, 空间表征, 空间更新