Intelligent system for early diagnosis of paediatric cerebral malaria from an electroencephalogram

Abstract

Background: This work is a part of a new concept. The aim is to study the measurements that can be used as input to an intelligent system for detecting cerebral malaria from an electroencephalogram (EEG).Materials and Methods: The study of brain connectivity through the calculation of the phase lag index, which is an adjacency matrix, allows evaluating the units such as degree, density and strength on each channel. These units were evaluated on 29 EEG recordings, consisting of twenty people suffering from coma and nine healthy individuals.Results: Considering analysis of variance with two factors that are frequency band and group (patient and healthy control), the degree and the density are always higher in healthy children compared to sick children. Nevertheless, the strength is always higher in healthy children compared to sick children with the exception of the delta band on which the values are equal and the alpha band on which the strength is higher in sick children by the report to healthy children.Conclusion: There is a significant difference (P = 0.002) between the strength of sick people compared to healthy people. Such technology could help reduce the death rate from malaria, in general, and cerebral malaria, in particular, especially in sub-Saharan Africa where this rate is very high.