眼跳（saccade）对于人与非人灵长类（如猕猴）来说是至关重要的，因为我们需要依赖于频繁的眼跳来获取外界精细的视觉信息。并且眼跳不是随机发生的，而是受一些相关脑区的控制，不同脑区在控制眼跳发生中所起的作用有所差异，因此可以通过检测并分析眼跳的方式来客观、快速地反映出大脑的功能。上述脑区的病变或损伤会导致异常眼跳的产生，比如可能会产生多步眼跳（multiple step saccades，MSS）。当受试者被要求做一步从注视点到目标位置的眼跳，多数情况下，受试者可以一步到位（single saccade），或者先进行一步大幅度的眼跳然后跟随一步朝向目标位置的小幅度纠正眼跳（corrective saccade，CS）。这两种眼跳形式被认为是正常地生理表现。但是还有一种眼跳形式，由两步及以上小幅度眼跳构成——多步眼跳。
这种眼跳的发生率在神经系统发育不成熟的幼年受试者以及一些患有神经退行性疾病的受试者中，尤其是在帕金森疾病（Parkinson’s Disease，PD）的患者中较高，因此被认为一种非生理表现。接下来，本文将围绕MSS在PD辅助诊断中的临床应用及其产生的神经机制两部分展开详细的介绍。
研究一：多步眼跳在帕金森疾病辅助诊断中的临床应用
众所周知，PD是发病率排在第二位的神经退行性疾病。年龄越高，患病几率越大；并且男性PD患者明显多于女性。虽然前人研究发现在记忆眼跳任务中提取分析MSS可以作为PD诊断的生物标记物，但是记忆眼跳任务较为复杂、难度较大，很多PD患者对于记忆眼跳范式指导语的理解和眼跳任务的执行都存在困难，因此极大的限制了MSS在PD临床诊断中的应用。此外，前人研究涉及的PD患者都为疾病中后期，截至目前，缺少MSS发生率在早期PD患者中的相关研究，如果通过检测MSS可以辅助诊断早期PD，从而提早干预的话，对于患者及其家庭乃至整个社会，都将是一个非常重大的福音。基于上述背景，提出研究问题如下：是否可以在较为简单的反射性眼跳中提取分析MSS来帮助诊断PD尤其是早期PD？为了回答上述问题，本文以PD病人、临床前期单侧PD模型猕猴（由单侧颈内动脉注射一种神经毒素1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine，MPTP，它可以特异性的杀伤黑质密布内的多巴胺神经元，因此可以帮助构建PD模型）以及对照组（包括健康与患有其他疾病的受试者，健康猕猴）作为研究对象开展了实验1、2，我们发现PD病人与猕猴的MSS发生率都显著地高于其他对照组，该结果表明在简单的、无创的、只需3-5分钟即可完成的反射性眼跳任务范式中提取分析MSS，可以为PD辅助诊断提供重要的参考。此外，由于年龄与性别是影响PD发病率的两大因素，为此研究一还探究了年龄与性别对MSS的影响（实验3），结果表明MSS的发生率与性别无关，与年龄（老化）呈正相关。
研究二：多步眼跳的神经机制
前人研究认为MSS的产生可能是由于眼跳计划过于自动（automatic）导致的，因为前人发现包含MSS的试次中的潜伏期明显短于一步到位试次中的潜伏期，并且通过经颅磁刺激（transcranial magnetic stimulation，TMS）干预与眼跳计划密切相关的两个脑区，即额叶眼跳区（frontal eye field， FEF）和额叶辅助眼跳区（supplementary eye field， SEF）可以增加MSS的发生率。但是这一理论仅仅依据一种视觉诱导的反射性眼跳任务，该理论是否具有普适性，截至目前未得到检验。因此本文提出研究二的问题：MSS产生是由于“眼跳计划过于自动”的理论是否具有普适性？为了回答上述问题，本文以健康受试者作为研究对象，以正向眼跳（pro-saccade）、反向眼跳（anti-saccade）、记忆眼跳（memory guided saccade）三个眼跳任务为行为范式开展实验4，发现“眼跳计划过于自动”的理论不具有普适性。因此，本文提出了一个关于MSS产生的一个理论假设，即MSS的产生是为了纠正眼跳落点错误而由大脑内部眼跳矫正模型触发的，它的产生受多个与眼跳相关的高级脑区调控。为了验证研究二提出的理论假设是否正确，本文以人类与猕猴为实验对象开展了实验5-7。具体来说，实验5：采用TMS在眼跳行为任务执行过程中干预前额叶脑区（prefrontal cortex，PFC）和后顶叶脑区（posterior parietal cortex，PPC）；实验6：进行眼跳行为任务态下的脑电活动数据同步采集。实验5-6的结果发现TMS刺激PFC与PPC后MSS发生率显著增加，并且在眼跳结束后的阶段，包含MSS的试次中PFC与PPC的gamma频段能量较高、PFC与PPC之间的格兰杰因果系数较大，上述结果表明PFC与PPC可能都参与调控MSS，但是PFC的调控作用要比PPC的调控作用大。此外，本文还以猕猴为研究对象，通过使临床前期单侧PD猕猴服用左旋多巴药物（实验7），发现猕猴服药后MSS发生率增加，但是猕猴的空间误差显著地减少了，说明MSS的产生受基底神经节的调控。因此，实验5-7的结果支持本文在实验4中提出的理论假设，即MSS的产生是由大脑内部眼跳矫正模型触发的，它的产生受多个与眼跳相关的高级脑区调控。
总的来说，全文聚焦于MSS，通过开展研究一与研究二，即实验1-7，揭示了简单的反射性眼跳任务中的MSS在PD早期辅助诊断中的临床意义，同时提出并验证了MSS产生的理论假设，即MSS是由大脑内部眼跳错误矫模型触发的，它的产生可能受多个与眼跳相关的高级脑区调控，该结果对推广MSS在PD辅助诊断中的临床应用和了解MSS产生的神经机制均有重要意义。
 

Saccades are crucial for humans and non-human primates (such as macaques) because we rely on frequent saccades to obtain detailed visual information from the external world. Saccades don’t occur randomly, but rather controlled by certain relevant brain regions, and different brain regions play various roles in the control of saccades. Therefore, detecting and analyzing saccades can reflect brain function objectively and quickly. Lesions or damage to the above brain regions can lead to abnormal saccades, such as the occurrence of multiple step saccades (MSS). When a subject is instructed to make a saccade from a fixation point to a target location, in most cases, the subject can achieve it in one step (single saccade), or make a large saccade followed by a small corrective saccade (CS) towards the target location. These two forms of saccades are treated as a normal physiological behavior. However, there is another form of saccade, consisting of two or more small saccades - multiple step saccades. The incidence of this type of saccade is particularly high in pediatric subjects with immature nervous system and in some patients with neurodegenerative diseases, especially in patients with Parkinson's Disease (PD). Therefore, it is considered a non-physiological performance. This article will provide a detailed introduction to the clinical application of MSS in the complementary diagnosis of PD and its neural mechanisms as following.    Study 1: The clinical application of multiple step saccades in the complementary diagnosis of Parkinson's disease
    As is known to all, PD is a neurodegenerative disease with the second highest incidence. The older the age, the higher the likelihood of developing the disease. In addition, male patients are significantly more common than female patients. Although previous studies have found that the extraction and analysis of MSS in memory guided saccade task can serve as a biomarker for PD diagnosis, this task is complex and difficult, and many PD patients have difficulty understanding the instruction and executing the task, which greatly limits the clinical application of MSS in PD diagnosis. In addition, the PD patients involving in the previous studies were in the middle and late stages of the disease, and so far, there is a lack of related research on the incidence of MSS in the early diagnosis of PD. If MSS could assist in the diagnosis of early PD so as to early intervention, it will be of great benefit for patients, their families and even the entire society. According to the above background, the research question is proposed as follows: Could MSS be extracted and analyzed in the simple reflexive saccade task to assist in PD diagnosis, especially in early PD? To answer the above question, we conduct experiments 1 and 2 by recruiting PD patients and preclinical hemi-PD model monkeys through unilateral internal carotid artery injection of MPTP, i.e., one type of neurotoxin that selectively kills dopamine neurons in the substantia nigra pars compacta, and control groups (including healthy subjects and patients with other diseases). We found that the incidence of MSS in such simple, non-invasive, and three to five minutes’ reflexive saccade task paradigm could provide an important reference for auxiliary diagnosis of PD. In addition, since age and gender are two of the major factors affecting the incidence of PD, we further explored the effect of age and gender on MSS (experiment 3), and we found that MSS is not related to gender and is positively correlated with age (aging).
Study 2: The neural mechanism of multiple step saccades
According to the findings of the previous studies, the generation of MSS might be induced by automatic saccadic plan, as the saccadic latency in trials including MSS was significantly shorter than that in trials with single saccade. Moreover, transcranial magnetic stimulation (TMS) interventions on the frontal eye field (FEF) and supplementary eye field (SEF), which are closely related to saccadic plan, increased the occurrence of MSS. However, this theory was only based on a visually guided reflexive saccade task, and its generalizability has not been tested so far. Therefore, in this article, we proposed research question 1 for Study 2: Does the "automatic saccadic plan" theory apply universally to the generation of MSS? To answer this question, we conducted Experiment 4 recruiting healthy participants and employing three different saccadic tasks, including pro-saccade, anti-saccade, and memory-guided saccade tasks. We found that the "automatic saccadic plan" theory was not universal. Therefore, we proposed an alternative theory for the generation of MSS, that is, MSS is triggered by an internal correction model to rectify saccadic spatial errors, which may be regulated by multiple saccade-related higher-level brain regions. To verify our theory, we conducted Experiments 5-7 by recruiting humans and monkeys. Specifically, in Experiment 5, we used TMS to intervene the prefrontal cortex (PFC) and posterior parietal cortex (PPC) of humans during saccadic tasks, and in Experiment 6, we recorded EEG signals of humans during saccadic tasks. Regarding to the results of experiments 5-6, we found that both PFC and PPC might be involved in the generation of MSS, whereas the PFC might play a more important role than PPC. Meanwhile, we study the mechanism of MSS by administering levodopa to the preclinical hemi-PD monkeys (experiments 7). We found that the incidence of MSS increased after taking L-dopa, but the spatial errors significantly reduced, indicating that the generation of MSS might be regulated by basal ganglia. Thus, the results of Experiments 5-7 supported our theory proposed in Experiment 4, which suggests that the generation of MSS might be regulated by multiple saccade-related higher-level brain regions.
Overall, this article focuses on MSS by conducting Study I and Study II, i.e., experiments 1-7, which reveal the clinical significance of MSS in the simple reflexive saccade task for the complementary diagnosis of early PD, while proposing and validating the theoretical hypothesis that MSS is triggered by the internal saccade error correction, which might be regulated by multiple higher brain regions associated with saccades. Such results are important for promoting the clinical application of MSS in the diagnosis of PD and understanding the neural mechanism of MSS generation.