基于脑功能磁共振影像的视觉信息解读技术

Functional magnetic resonance imaging "brain reading"of visual information: a review

介绍了以脑信号视觉计算模型为重点的视觉信息编码技术和以脑信号分类与识别为重点的视觉信息解码技术的发展现状;重点阐述了编码研究中的Gabor感受野模型、形式相似性分析方法和解码研究中的多体素分析方法,并讨论了分类器设计中的线性和非线性分类器以及特征选择问题;最后对该技术的发展趋势进行了展望。

This paper first reviewed the encoding teniques focused on computational models and decoding teniques focused on classification and identification,particullarly the Gabor receptive-field model,representational similarity analysis method in encoding and the multivariate pattern analysis（MVPA） approach in decoding.In addition,it also reviewed the linear and nonlinear classifiers and feature selection approaches in decoding models.In the end,it suggested future research directions of ＂brain reading＂ with fMRI.