Spatial Localization in Target Detection Based on Decoding N2pc Component.

Yijing Wang,Zhiguo Luo,Shaokai Zhao,Liang Xie,Minpeng Xu,Dong Ming,Erwei Yin
DOI: https://doi.org/10.1016/j.jneumeth.2021.109440
IF: 2.987
2022-01-01
Journal of Neuroscience Methods
Abstract:BACKGROUND:The Gaze-independent BCI system is used to restore communication in patients with eye movement disorders. One available control mechanism is the utilization of spatial attention. However, spatial information is mostly used to simply answer the "True/False" target recognition question and is seldom used to improve the efficiency of target detection. Therefore, it is necessary to utilize the potential advantages of spatial attention to improving the target detection efficiency.NEW METHOD:We found that N2pc could be used to assess spatial attention shift and determine target position. It was a negative wave in the posterior brain on the contralateral target stimulus. From this, we designed a novel spatial coding paradigm to achieve two main purposes at each stimulus presentation: target recognition and spatial localization.COMPARISON WITH EXISTING METHODS:We used a two-step classification framework to decode the P300 and N2pc components.RESULTS:The average decoding accuracy of fourteen subjects was 84.43% (σ = 1.14%), and the classification accuracy of six subjects was more than 85%. The information transfer rate of the spatial coding paradigm could reach 60.52 bits/min. Compared with the single stimulus paradigm, the target detection efficiency was successfully improved by approximately 10%.CONCLUSIONS:The spatial coding paradigm proposed in this paper answered both "True/False" and "Left/Right" questions by decoding spatial attention information. This method could significantly improve image detection efficiencies, such as visual search tasks, Internet image screening, or military target determination.
What problem does this paper attempt to address?