军事医学2025，Vol.49Issue(11)：812-822,11.DOI:10.7644/j.issn.1674-9960.2025.11.002
视觉感知与想象下的fNIRS比较与分析研究
fNIRS responses during visual perception and imagery:a comparative analysis
摘要
Abstract
Objective To compare brain region activation,hemodynamic responses,and functional connectivity between visual perception and visual imagery using functional near-infrared spectroscopy(fNIRS),and to evaluate the performance of different methods of feature extraction and classification in recognizing visual cognitive states.Methods Hemodynamic signals were recorded from 34 healthy participants using the Shimadzu LIGHTNIRS system during tasks involving the visual perception and imagery of 340 images.Statistical analysis and machine learning approaches were employed to explore the neural representations underlying these processes.Results Analysis of changes inoxygenated hemoglobin(HbO)concentrations suggested that visual perception predominantly activated the occipital regions,with higher response amplitudes and a shorter rise latency.In contrast,visual imagery mainly involved prefrontal activation,with a longer rise latency.General linear model(GLM)analysis showed that the way prefrontal and occipital cortices were activated was considerably different between the two tasks.Functional connectivity analysis indicated enhanced prefrontal-occipital connections during imagery despite the lower overall network efficiency in imagination.Classification results indicated that the support vector machine(SVM),adaptive boosting(AdaBoost),and extreme gradient boosting(XGBoost)could all clearly distinguish the two tasks,with XGBoost showing the highest accuracy of 89.1%.Feature ablation analysis identified the skewness,peaklatency,and slope as the best discriminative features.Conclusion Visual perception and imagery show systematic differences in spatial activation,temporal response,and network connectivity.By integrating fNIRS with multidimensional feature modeling and machine learning approaches,the cognitive state can be recognized efficiently.关键词
功能性近红外光谱技术(fNIRS)/视觉感知/视觉想象/血氧变化/功能连接/模式识别Key words
functional near-infrared spectroscopy(fNIRS)/visual perception/visual imagery/hemodynamic response/functional connectivity/pattern recognition分类
医药卫生引用本文复制引用
李伟龙,樊桦,赵佳欣,贾淑钰,杜博爱,何振,应晓敏..视觉感知与想象下的fNIRS比较与分析研究[J].军事医学,2025,49(11):812-822,11.基金项目
国家重点研发计划项目(2022YFF1202400) （2022YFF1202400）
