Parameters of stochastic models for electroencephalogram data as biomarkers for child’s neurodevelopment after cerebral malaria

Abstract

The objective of this study was to test statistical features from the electroencephalogram (EEG) recordings as predictors of neurodevelopment and cognition of Ugandan children after coma due to cerebral malaria. The increments of the frequency bands of EEG time series were modeled as Student processes; the parameters of these Student processes were estimated and used along with clinical and demographic data in a machine-learning algorithm for the prediction of children’s neurodevelopmental and cognitive scores 6 months after cerebral malaria illness. The key innovation of this work is in the identification of stochastic EEG features that can serve as language-independent markers of the impact of cerebral malaria on the developing brain. The results can enhance prognostic determination of which children are in most need of rehabilitative interventions, which is especially important in resource-constrained settings such as sub-Saharan Africa.

Artificial intelligence and the future of global health

“…40 Researchers have used this approach to quantify the risk of tuberculosis treatment failure 41 and assess the risk of cognitive sequelae after malaria infection in children. 42 See Online for appendix…”

Generalized Gaussian time series model for increments of EEG data

“…Since the heavy-tailed subfamily of GGD (1) for s = 2 comes down to the Student-type distribution, in this case the AR (1) process constructed in such a way could be used as a time-series model for EEG signals. This AR(1) time series is used for modeling the EEG signals in [40]. Furthermore, in light-tailed (b = 0) case of the GGD (1) for s ∈ (0, 1] ∪ {2} the strictly-stationary AR(1) time series with GGD (1) marginals could be constructed since for these values of parameter s this distribution is infinitelydivisible and self-decomposable, see [11,Theorems 5 and 6].…”

“…Data used in this analysis were used previously in [40]. Data were collected during the observational study of the pathogenesis of severe malaria (cerebral malaria and severe malarial anemia) in surviving children.…”

“…An overview of these models is given in [36]. Statistical analyses of EEG data included classification and prediction using arrays of EEG features, but few models for the underlying stochastic processes have been proposed [33,40]. In this paper we build upon past work [40] in which it was determined that the underlying stochastic processes across channels are not stationary but have stationary increments.…”

“…Statistical analyses of EEG data included classification and prediction using arrays of EEG features, but few models for the underlying stochastic processes have been proposed [33,40]. In this paper we build upon past work [40] in which it was determined that the underlying stochastic processes across channels are not stationary but have stationary increments. The distributions of these increments were symmetrical across channels; in some channels the distributions were heavy-tailed whereas in other channels, lighter tails were observed.…”