不同眼病前额脑电响应的共性及视力的 中介作用

Commonality of Prefrontal EEG Responses in Different Eye Diseases and the Mediating Role of Visual Acuity

摘要: 目的：探讨不同解剖部位眼病(屈光不正、晶状体浑浊、眼底疾病)状态下，单通道前额事件相关电位(ERP)响应的共性与特异性，验证前额脑电特征能否用于眼病的鉴别诊断，并明确“视敏锐度”在视觉信息皮层重构中的中介调节作用。方法：招募50例(71只患眼)确诊为单纯屈光不正、白内障及眼底疾病的受试者。在虚拟现实(VR)视觉刺激下同步采集单通道前额脑电，提取P1、P2、P3、N1、N2特征成分。采用协方差分析(ANCOVA)排除年龄及基础视力混杂因素，通过多元Logistic回归及判别分析评估疾病分类模型的准确率，并利用Pearson/Spearman相关性分析评估ERP特征与Log MAR视力的关系。结果：在排除年龄与视力协变量影响后，三组间前额ERP各项特征成分的潜伏期、峰谷值及响应面积差异均无统计学意义(P > 0.05)。基于前额ERP特征构建的多分类判别分析模型对三种眼病的分类准确率极低(仅13.33%~33.33%)。相反，ERP中N1潜伏期和N2响应面积与Log MAR视力呈显著正相关(P < 0.05)。结论：不同致病机制的眼病在向高级认知脑区传递信息时表现出高度的电生理共性。前额ERP对眼病类型缺乏鉴别特异性，但对视觉功能损伤程度高度敏感。视敏锐度是驱动前额认知资源调用的核心中介变量，本研究从神经电生理维度验证了视觉损害的“共同通路原则”，为开发便携式客观视功能评估设备提供了循证基础。

Abstract: Purpose: The differential diagnosis of ocular diseases involving different anatomical sites remains a clinical challenge, especially when objective assessment of visual function is required. This study aimed to explore the commonalities and specificities of single-channel prefrontal Event-Related Potential (ERP) responses in patients with three distinct types of ocular diseases that affect different anatomical sites, namely refractive errors, lens opacities, and fundus diseases. Specifically, we sought to verify the potential value of prefrontal Electroencephalographic (EEG) features in the differential diagnosis of these ocular diseases and to clearly define the mediating and regulatory role of visual acuity in the cortical reconstruction of visual information, which could provide new insights into the neuroelectrophysiological mechanisms underlying visual impairment. Methods: A total of 50 subjects (71 affected eyes) who were clinically diagnosed with simple refractive errors, cataracts (as a representative of lens opacities), and fundus diseases were enrolled in this study. All subjects underwent a comprehensive ophthalmic examination to confirm the diagnosis and exclude other concurrent ocular or systemic diseases that might affect the results. During the experiment, single-channel prefrontal EEG signals were synchronously collected under standardized Virtual Reality (VR) visual stimulation, which was designed to ensure consistent and controllable visual input. The key characteristic components of ERP, including P1, P2, P3, N1, and N2, were extracted and analyzed. To eliminate the potential confounding effects of age and baseline visual acuity on ERP results, Analysis of Covariance (ANCOVA) was applied. Subsequently, multivariate Logistic regression and discriminant analysis were employed to evaluate the accuracy of the disease classification model constructed based on the extracted prefrontal ERP features. Additionally, Pearson correlation analysis (for normally distributed data) and Spearman correlation analysis (for non-normally distributed data) were used to explore the association between ERP features and Log MAR visual acuity, which is a standard indicator for evaluating visual function. Results: After adjusting for the covariates of age and visual acuity, statistical analysis showed that there were no significant differences in the latency, peak amplitude, and response area of each prefrontal ERP component (P1, P2, P3, N1, N2) among the three groups of patients with different ocular diseases (all P > 0.05). The multi-class discriminant model constructed based on the prefrontal ERP features exhibited an extremely low classification accuracy for the three types of ocular diseases, with the accuracy ranging only from 13.33% to 33.33%, indicating that prefrontal ERP features are not effective for differentiating these ocular disease types. In contrast, the results of correlation analysis revealed that N1 latency and N2 response area in ERP were significantly positively correlated with Log MAR visual acuity (all P < 0.05), suggesting that these ERP components are closely associated with the severity of visual function impairment. Conclusion: Ocular diseases with different pathogenic mechanisms, which affect different anatomical sites of the eye, exhibit high electrophysiological commonalities when transmitting visual information to high-order cognitive brain regions. Prefrontal ERP lacks specificity in differentiating the types of ocular diseases but is highly sensitive to the degree of visual function impairment, which means it can effectively reflect the severity of visual damage. Visual acuity serves as a core mediating variable that drives the allocation of prefrontal cognitive resources during visual information processing. From the neuroelectrophysiological perspective, this study verifies the “common pathway principle” of visual impairment, which holds that different types of visual damage may converge on a common neural pathway in the brain. Furthermore, these findings provide important evidence-based support for the development of portable, objective visual function assessment devices, which could be particularly useful in clinical settings where subjective visual assessment is not feasible. Visual stimulation, and characteristic components (P1, P2, P3, N1, N2) were extracted. Analysis of Covariance (ANCOVA) was used to exclude confounding factors of age and baseline visual acuity. Multivariate Logistic regression and discriminant analysis were employed to evaluate the accuracy of the disease classification model, and Pearson/Spearman correlation analysis was used to explore the association between ERP features and Log MAR visual acuity. Results: After adjusting for age and visual acuity covariates, there were no significant differences in the latency, peak amplitude, and response area of each prefrontal ERP component among the three groups (P > 0.05). The multi-class discriminant model constructed based on prefrontal ERP features had an extremely low classification accuracy of only 13.33%~33.33% for the three types of ocular diseases. In contrast, N1 latency and N2 response area in ERP were significantly positively correlated with Log MAR visual acuity (P < 0.05). Conclusion: Ocular diseases with different pathogenic mechanisms exhibit high electrophysiological commonalities when transmitting information to high-order cognitive brain regions. Prefrontal ERP lacks specificity in differentiating ocular disease types but is highly sensitive to the degree of visual function impairment. Visual acuity is a core mediating variable driving the allocation of prefrontal cognitive resources. This study verifies the “common pathway principle” of visual impairment from the neuroelectrophysiological perspective and provides evidence-based support for the development of portable objective visual function assessment devices.