Real-time nerve decoding system based on brain function features

Abstract

The invention provides a real-time nerve decoding system based on brain function features. A real-time functional magnetic resonance imaging technology is used, the brain function features which are extracted on line are classified, and real-time nerve decoding is conducted on magnetic resonance functional images. The real-time nerve decoding system comprises a preprocessing module, a feature extraction module, a classification decoding module, a display and feedback module and a parameter configuration module. After the preprocessing module conducts on-line reading and format conversion on the magnetic resonance functional images, the signal to noise ratio of the images is improved through real-time head motion correction, spatial normalization and smoothness; then the feature extraction module is used for extracting the main ingredients of whole brain image signals to be used as the features; the classification decoding module conducts classification decoding on the features through a real-time support vector machine or a Gaussian process classifier, and the display and feedback module feeds the classification accuracy rate back to an examinee in real time. The real-time nerve decoding system based on the brain function features is of great application value to analyzing psychological traits, recognizing brain activities, improving cognitive functions and the like.

Description

technical field [0001] The invention provides a real-time neural decoding system based on brain function characteristics, which belongs to the technical field of information processing, and specifically refers to extracting brain function characteristics by analyzing real-time functional magnetic resonance imaging data, and performing real-time feature classification on the characteristics, thereby obtaining brain information. Cognitive state; at the same time, subjects can observe their own classification results online, and adjust the activity state of their brains through corresponding strategies, so as to improve the corresponding cognitive functions. Background technique [0002] Neural decoding technology refers to the use of brain activity signals to identify stimulus information by establishing a direct mapping relationship from brain activity to stimulus. Currently, the signals used in neural decoding mainly include neural electrophysiological signals, such as EEG, ...

AI Technical Summary

Problems solved by technology

However, real-time neural decoding is still blank in China. The existing decoding system is still offline and cannot meet the requirements of neurofeedback; and most of the feature extraction methods used are based on the region of interest, while the classification method is mainly based on the support vector Machine-based, the optimization process of this classifier is more complicated

Images