Diagnosis of autism through EEG processed by advanced computational algorithms: a pilot study

Highlights • Autism is a devastating disease affecting 1-2% of newborns. Its incidence is steadily increasing. • A growing body of evidence points out that to effectively treat autism, the earliest possible detection is directly proportional to successful therapy. This clearly implies that the availability of an accurate and relatively inexpensive diagnostic method for early diagnosis should be one of the medical community's highest priorities. • The relevant involvement of the cerebral cortex in substantially altering cortical circuitry explains the unique pattern of deficits and strengths that characterize cognitive functioning. On the other, this makes the potential usefulness of EEG recording plausible as a biomarker of these abnormalities. • Despite this plausibility, very few studies have attempted to use EEG recording in diagnosing autism. • Traditional EEG measures apply frequency domain analysis, assume that EEG is stationary and employ linear feature extraction. Novel approaches based on artificial adaptive systems like those employed by us, apply data driven time& frequency domain analysis, assume that EEG is not stationary and use nonlinear feature extraction. • To our knowledge, this is the first study that applies an artificial adaptive system to extract interesting features in computerized EEG that distinguishes ASD children from typically developing ones. • The results are extremely interesting and open new avenues for confirmatory clinical studies in this important field.