Unilateral spatial neglect (USN) is a common post-stroke cognitive impairment that severely affects the functional recovery and quality of life of patients. Therefore, objective assessment of USN is essential to facilitate patients' functional recovery.

To investigate the feasibility of using adynamic eye-tracking task to assess USN in sub acute stroke patients.

Thirty inpatients in the sub acute phase after stroke in Department of Rehabilitation Medicine, Nanchong Central Hospital were recruited from September 2021 to July 2022. USN was assessed using the Behavioral Inattention Test-Conventional (BIT-C), Catherine Bergego Scale (CBS), and a two-minute dynamic eye-tracking task within one week of admission by two professional rehabilitation therapists, and adverse effects during the assessment were recorded. USN was diagnosed by a CBS score >0, a total BIT-C score <129, or gaze points distributed outside the four screen zones (gaze points distributed within the four screen zones were defined as non-USN). Correlation and consistency analyses were used to evaluate the results of the three assessments.

In accordance with the performance assessment in conducting the dynamic eye-tracking task, 14 patients had left-sided neglect, and the other 16 had no USN. USN patients had a higher percentage of gaze points on the right side of the screen than non-USN patients (Z=-4.776, P<0.001). There was a significant difference in percentages of gaze points on the left and right side of the screen in USN patients (Z=-3.49, P<0.001). By BIT-C assessment, 15 patients were diagnosed with USN, and the remaining 15 patients had no USN. CBS assessment revealed that 16 patients had varying degrees of USN, and the remaining 14 were non-USN patients. The BIT-C and dynamic eye-tracking task were highly consistent in the assessment of USN (Kappa=0.933, P<0.001). Spearman's rank correlation analysis revealed a negative correlation between the percentage of gaze points on the right side of the screen and the total BIT-C score (r_s=-0.776, P<0.001). The CBS and dynamic eye-tracking task had good agreement on the assessment of USN (Kappa=0.867, P<0.001) .

It is feasible to use the dynamic eye-tracking task to assess USN, because it is less time-consuming, with high level of patient participation and motivation, and there is a good consistency between the results assessed by it and BIT-C or CBS. Therefore, it can be used as a supplement to the standard USN assessment.

单侧空间忽略（USN）是脑卒中后常见的认知障碍之一，严重影响患者的功能恢复和生活质量。因此，客观地评估对于促进患者功能恢复至关重要。

探究基于眼动追踪的动态任务评估亚急性期脑卒中患者USN的可行性。

选取2021年9月—2022年7月在南充市中心医院康复医学科招募30例亚急性期脑卒中住院患者为研究对象。采用行为不集中测试的常规分项（BIT-C）、凯瑟琳波哥量表（CBS）以及基于眼动追踪的动态任务评估患者USN情况，入院1周内完成，并记录评估过程中的不良反应。CBS得分>0分被诊断为USN。BIT-C总分<129分被诊断为USN。基于眼动追踪的动态任务评估过程共计2 min，根据患者注视点的空间分布情况将患者分为：非USN患者（注视点分布于4个区域）和USN患者（注视点未分布于4个区域），由2名专业的康复治疗师完成患者的临床量表评估和基于眼动追踪的动态任务评估。对3项评估结果进行相关性分析和一致性分析。

基于眼动追踪的动态任务评估结果显示，30例患者中有14例患者为左侧USN患者，16例患者为非USN患者。USN患者在右侧屏幕注视点占比高于非USN患者（Z=-4.776，P<0.001）；USN患者左右侧屏幕间注视点占比比较，差异有统计学意义（Z=-3.490，P<0.001）。BIT-C评估结果显示，15例患者BIT-C总分<129分，被诊断为USN患者，余下15例为非USN患者。CBS评估结果显示，16例患者存在不同程度的USN，其余14例为非USN患者。BIT-C和基于眼动追踪的动态任务在评估USN患者的检出情况上高度一致（Kappa=0.933，P<0.001）；Spearman秩相关分析结果显示，患者右侧注视点占比与BIT-C总分呈负相关（r_s=-0.776，P<0.001）。CBS和基于眼动追踪的动态任务在评估USN患者的检出情况上有较高的一致性（Kappa=0.867，P<0.001）。

基于眼动追踪的动态任务评估脑卒中后USN是可行的，其评估结果与BIT-C和CBS评估结果具有较高的一致性，且评估过程耗时较少，患者参与度高、积极性强，可以作为标准USN评估的一种补充。