EEG-Based Functional Connectivity during Progression from Infantile Spasms to Lennox Gastaut Syndrome

Abstract

Infantile Epileptic Spasms Syndrome (IESS) is a severe infantile epilepsy that can progress into Lennox-Gastaut Syndrome (LGS), associated with intellectual problems, and psychiatric disorders. Early diagnosis and treatment of LGS may improve prognosis [1]. There is a need for biomarkers to monitor the progression of these children. Based on prior work [2], [3], we hypothesize that functional connectivity strength is a robust biomarker for the presence of these epilepsies. Five children diagnosed with IESS who progressed to LGS were included in this study. Functional connectivity networks were obtained by performing the statistical analysis of cross-correlation between electrode pairs [2]. The number of strong connections and the mean strength of the top 10% of connections were correlated to the disease state progression, response to treatment, and the child’s age at the time of the EEG. The number of strong connections and the mean connection strength gave approximately equivalent results. The connectivity strength was high at the time of IS and LGS diagnosis. Positive treatment outcome was associated with a decrease in strength, while an increase in strength reflected a worsening of the disease. Further, connectivity strength was not correlated to age, suggesting that these network changes are not due to age-related physiological changes. Functional connectivity strength reflected the presence of IS and LGS, as well as positive or negative response to treatment. Computational EEG analysis of functional connectivity could be applied in clinical practice to improve the prognosis of LGS patients. However, it is critical to extend this analysis to a larger cohort of subjects to increase its power and validate these results.