Services on Demand
Journal
Article
Indicators
- Cited by SciELO
- Access statistics
Related links
- Similars in SciELO
Share
RISTI - Revista Ibérica de Sistemas e Tecnologias de Informação
Print version ISSN 1646-9895
Abstract
ESTEBAN, Velásquez R.; ALEJANDRO, Cardona M. and ALEJANDRO, Peña P.. Vector Model for the Inference of the Cognitive State in Derived State Coma Patients. RISTI [online]. 2014, n.13, pp.65-81. ISSN 1646-9895. https://doi.org/10.4304/risti.13.65-81.
The traditional method to evaluate the conscious state of an individual consists in applying a stimulus and analyzing the response. However this technique is limited when the person cannot response evidentially to that stimulus, as are derived state coma patients. In such cases, a direct connection to the brain is required to detect the response. Therefore, in this paper we develop and analyze a computational model employing support vector machines (SVM) to infer the cognitive state of derived state coma patients using an affordable electroencephalography neuroheadset. The results given by the proposed model confirmed that the model can correctly classify the cognitive state in at least 4 out of 5 tests in control patients, which can be translated in the contribution of a low cost system for the analysis of the conscious state and the possible following communication with some derived state coma patients by defined medical protocols. Thus, this system is a contribution for clinics and hospitals as a potential diagnostic tool.
Keywords : Support vector machine (SVM); Polynomial kernel; Gaussian kernel; Electroencephalography (EEG); Brain-computer-interface (BCI); Vegetative state.