| Тойм: | <p>Autism Spectrum Disorders (ASD) are a group of lifelong neurodevelopmental disorders, described by three core deficits which are thought to be consequences of atypical cortical activity. Abnormal neural connectivity underlying ASD could be inferred from dynamical system characteristics of the brain measured from electroencephalography (EEG) time series. Simple EEG measurement have the potential to provide important clinical biomarkers for early identification, risk assessment and monitoring the progression of ASD, while spanning the spectrum’s heterogeneity of severity. The explicit goal of this project was to use Multiscale Entropy (sample entropy) measures and Recurrence Quantitative Analysis (RQA) on EEG to identify quantifiable neural correlates of behaviours associated with an ASD diagnosis. The hypotheses were tested on two cohorts: the Kenyan cohort with data collected in Kilifi, Kenya from both neurotypical children as well as children diagnosed with ASD and the NDAR cohort from the National Database for Autism Research (NDAR), also containing data from both groups. The results showed that complexity measured using sample entropy is useful in distinguishing the two groups, both at a whole brain level in the alpha, beta and gamma bands, and at a single electrode level, mainly in the theta band in the frontal, occipital and parietal electrodes and in the high gamma band in the frontal and prefrontal areas. RQA variables analysis showed that determinism can delineate ASD in the theta, gamma and high gamma bands in all electrode locations, in the delta band in the left frontal and temporal areas and in the alpha band in the left hemisphere in the occipital, parietal, temporal and central electrodes; laminarity is useful in delineating ASD in the frontal, temporal and parietal regions in the theta band and the occipital, parietal and temporal regions in the delta band; l_entropy played a role in ASD delineation in the central, temporal and occipital areas in the theta frequency; lastly, l_max accurately distinguished the two groups, mainly in the frontal and central areas in the theta band and in the parietal region in the alpha frequency. These findings represent pilot evidence of potential high utility of this method, which can have great impact on clinical practice, in the early screening and diagnosis stages of ASD.</p>
|
|---|