Abstracts

Rationale: Autonomic cardiac dysfuction is reported in patients with TLE, manifesting as heart rate abnormalities. However, reports measuring HR using frequency-domain analysis (fourier transformation) have not controlled for respiratory arrhythmia and yielded conflicting results. As biological systems normally show characteristics of variability which can be mathematically quantified as non-linear entropy, we hypothesize that by measuring the entropy of cardiac rhythms, we can distinguish TLE patients from those with non-epileptic seizures. Methods: We retrospectively identified (2003-2008) two groups of patients from our epilepsy monitoring unit (EMU). All patients with diagnosed cardiovascular morbidities were excluded. Our experimental group consisted of patients with confirmed right TLE through a seizure free outcome after temporal lobectomy, and our controls of patients with electroencephalographically confirmed pseudoseizures. For each patient we extract three 120s periods (awake, sleep state and proceeding seizure onset) from EMU files that recorded EEG and EKG simultaneously. We used phase portraits to describe the oscillatory nature of cardiac dynamics, and quantify the coupling between cardiac and respiratory systems.Results: Five patients were included in the control pseudoseizure group (4 women; mean age 31 12years) and four patients in the right TLE group (3 women; mean age 32 15years). TLE patients had higher average entropy value (S = 4.5) than the pseudoseizure group (S = 2.5) in all three states. Calculation of the approximate entropy for each state of consciousness showed distinct peaks for each group.Conclusions: This simple noninvasive investigation of HR distinguishes epileptic from non-epileptic patients. The higher entropy values seen in TLE signify higher rates of HR fluctuations in comparison to autonomic 'healthy'patients. The dynamics of coupled phase oscillators provides the framework to characterize the nonlinear behavior associated with dysautonomia in patients with TLE.