Abstract: Objective  To explore the relationship between cognitive load and different physiological indicators by dynamic monitoring the physiological indicators in the motor imagery training in patients with acute ischemic stroke. 
Methods  A descriptive correlational research design was adopted, and patients with acute ischemic stroke who met the inclusion criteria were consecutively enrolled for assessments from February to November 2022 at the Department of Neurology of Xuanwu Hospital, Capital Medical University. Patients mimicked the grasping actions on the computer screen while holding a ball with both hands. They first performed left-hand grasping imagery training (4 seconds) three times, then rested for 3 seconds. After that, they performed right-hand grasping imagery training (4 seconds) three times, then rested for 3 seconds again. A total of 42 rounds and 252 grasping imagery training sessions were conducted, with a total duration of approximately 20 minutes. Before the start of motor imagery training and immediately after its completion, the cognitive load scale was used to measure the cognitive load, and the physiological indicators were monitored throughout the training. The relationship between cognitive load and physiological indicators was analyzed using latent profile and correlation analyses. 
Results  A total of 102 patients with acute ischemic stroke were included, with an average age of (58.94±13.07) years, the time from onset to treatment of 2 (1-5) days, and the NIHSS score of 3 (2-5) points. Results from latent profile and correlation analyses revealed  that the latent types of cognitive load in patients with acute ischemic stroke before motor imagery training were divided into three classes, specifically “low cognitive load in all dimensions type” “high effort and zero physical requirements type” and “high mental requirements-self-expression type”. Heart rate, pulse rate, respiratory rate, blood oxygen saturation, systolic blood pressure, diastolic blood pressure, coefficient of variability of systolic blood pressure, coefficient of variability of diastolic blood pressure, heart rate immediately after the end, pulse rate immediately after the end, respiratory rate immediately after the end, and blood oxygen saturation immediately after the end were correlated with the total score of cognitive load and various dimensions of cognitive load. The latent types of cognitive load during motor imagery training were divided into four classes, specifically “high time limit requirements-low frustration type” “high physical requirements-time limit requirements-effort level-frustration and low self-expression type” “high effort level and low self-expression-frustration type” and “high physical requirements-effort level-frustration and low self-expression type”. Heart rate, pulse rate, respiratory rate, blood oxygen saturation, systolic blood pressure, diastolic blood pressure, mean arterial pressure, coefficient of variability of systolic blood pressure, coefficient of variability of diastolic blood pressure, heart rate immediately after the end, pulse rate immediately after the end, respiratory rate immediately after the end, blood oxygen saturation immediately after the end, systolic blood pressure immediately after the end, diastolic blood pressure immediately after the end, and mean arterial pressure immediately after the end were correlated with the total score of cognitive load and various dimensions of cognitive load. The results of the canonical correlation analysis between cognitive load and vital signs showed that among classes one to four in motor imagery training, the first canonical correlation pair in class one was statistically significant (P<0.001), with a canonical correlation coefficient＞0.999. The absolute values of the standardized coefficients for systolic and diastolic blood pressure variability were relatively high.
Conclusions  In patients with acute ischemic stroke, there were correlations between cognitive load and vital signs across latent profile classes before and during motor imagery training. Notably, the coefficients of variability of systolic and diastolic blood pressure indicated cognitive load of patients in class one during training more accurately.

Key words: Acute ischemic stroke; Motor imagery; Cognitive load; Physiological indicator; Correlation analysis

摘要： 目的 通过对急性缺血性卒中患者运动想象训练中生理指标的动态监测，探究认知负荷与不同生理指标间的关系。 
方法 采用描述性相关研究设计，连续纳入2022年2—11月首都医科大学宣武医院神经内科符合纳入标准的急性缺血性卒中患者进行测评。患者在双手握球状态下模仿计算机屏幕中的抓握动作，先进行左手抓握想象训练（4 s）3次，休息3 s，再进行右手抓握想象训练（4 s）3次，休息3 s，共进行42轮、252次抓握想象训练，总时长约为20 min。在患者运动想象训练前、训练结束即刻，利用认知负荷量表测量认知负荷，训练全程监测患者生理指标，通过潜在剖面分析与相关性分析对患者认知负荷与生理指标的关系进行探究。
结果 共纳入102例急性缺血性卒中患者，平均年龄（58.94±13.07）岁，发病到就诊时间2（1～5）d，NIHSS评分3（2～5）分。潜在剖面分析与相关性分析结果显示，急性缺血性卒中患者运动想象训练前的认知负荷潜在分型分为3型，具体为“认知负荷全维度低谷型”“高努力程度与零体力要求型”“高脑力要求-自我表现型”。心率、脉率、呼吸频率、血氧饱和度、收缩压、舒张压、收缩压变异系数、舒张压变异系数、结束即刻心率、结束即刻脉率、结束即刻呼吸频率、结束即刻血氧饱和度与认知负荷总分、认知负荷各维度具有不同程度的相关性。运动想象训练过程中认知负荷的潜在分型分为4型，具体为“高时限要求-低受挫程度型”“高体力要求-时限要求-努力程度-受挫程度与低自我表现型”“高努力程度与低自我表现-受挫程度型”“高体力要求-努力程度-受挫程度与低自我表现型”。心率、脉率、呼吸频率、血氧饱和度、收缩压、舒张压、平均动脉压、收缩压变异系数、舒张压变异系数、结束即刻心率、结束即刻脉率、结束即刻呼吸频率、结束即刻血氧饱和度、结束即刻收缩压、结束即刻舒张压、结束即刻平均动脉压与认知负荷总分、认知负荷各维度具有不同程度的相关性。认知负荷与生命体征的典型相关分析结果显示，在运动想象训练中的类别1～4中，类别1的第1对典型相关对具有统计学意义（P＜0.001），典型相关系数＞0.999，其中收缩压变异系数、舒张压变异系数的标准化系数绝对值较大。
结论 在急性缺血性卒中患者运动想象训练前、训练过程中各认知负荷潜在剖面类别分组的认知负荷与生命体征存在相关性，其中，收缩压变异系数、舒张压变异系数能更精确地反映训练中类别1患者的认知负荷。

关键词: 急性缺血性卒中; 运动想象; 认知负荷; 生理指标; 相关性分析