Suppr超能文献

模拟人工视觉下人脸与非人脸识别早期视觉处理的神经生理学研究

Neurophysiology study of early visual processing of face and non-face recognition under simulated prosthetic vision.

作者信息

Yang Yuan, Guo Hong, Tong Shanbao, Zhu Yisheng, Qiu Yihong

机构信息

Department of Biomedical Engineering, Shanghai Jiao Tong University, China.

出版信息

Annu Int Conf IEEE Eng Med Biol Soc. 2009;2009:3952-5. doi: 10.1109/IEMBS.2009.5333672.

Abstract

Behavioral researches have shown that the visual function can be partly restored by phosphene-based prosthetic vision for the non-congenital blinds. However, the early visual processing mechanisms of phosphene object recognition is still unclear. This paper aimed to investigate the electro-neurophysiology underlying the phosphene face and non-face recognition. The modulations of latency and amplitude of N170 component in the event-related potential (ERP) were analyzed. Our preliminary results showed that (1) both normal and phosphene face stimuli could elicit prominent N170; nevertheless, phosphene stimuli caused notable latency delay and amplitude suppression on N170 compared with normal stimuli and (2) under phosphene non-face stimuli, a slight but significant latency delay occurred compared with normal stimuli, while amplitude suppression was not observed. Therefore, it was suggested that (1) phosphene perception caused a disruption of the early visual processing for non-canonical images of objects, which was more profound in phosphene face processing; (2) the face-specific processing was reserved under prosthetic vision and (3) holistic processing was the major stage in early visual processing of phosphene face recognition, while part-based processing was attenuated due to the loss of the details.

摘要

行为研究表明,基于光幻视的假体视觉可以部分恢复非先天性盲人的视觉功能。然而,光幻视物体识别的早期视觉处理机制仍不清楚。本文旨在研究光幻视面部和非面部识别背后的神经电生理学。分析了事件相关电位(ERP)中N170成分的潜伏期和波幅的调制情况。我们的初步结果表明:(1)正常面部刺激和光幻视面部刺激均可诱发明显的N170;然而,与正常刺激相比,光幻视刺激导致N170的潜伏期明显延迟和波幅抑制;(2)在光幻视非面部刺激下,与正常刺激相比,潜伏期出现轻微但显著的延迟,而未观察到波幅抑制。因此,有人提出:(1)光幻视感知导致对物体非典型图像的早期视觉处理中断,这在光幻视面部处理中更为明显;(2)在假体视觉下,面部特异性处理得以保留;(3)整体处理是光幻视面部识别早期视觉处理的主要阶段,而基于部分的处理由于细节缺失而减弱。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验