Epidemic models characterize seizure propagation and the effects of epilepsy surgery in individualized brain networks based on MEG and invasive EEG recordings

Abstract

Background Epilepsy surgery is the treatment of choice for drug-resistant epilepsy patients. However, seizure-freedom is currently achieved in only 2/3 of the patients after surgery. In this study we have developed an individualized computational model based on functional brain networks to explore seizure propagation and the efficacy of different virtual resections. Eventually, the goal is to obtain individualized models to optimize resection strategy and outcome. Methods We have modelled seizure propagation as an epidemic process using the susceptible-infected (SI) model on individual functional networks derived from presurgical MEG. We included 10 patients who had received epilepsy surgery and for whom the surgery outcome at least one year after surgery was known. The model parameters were tuned in order to reproduce the patient-specific seizure propagation patterns as recorded with invasive EEG. We defined a personalized search algorithm that combined structural and dynamical information to find resections that maximally decreased seizure propagation for a given resection size. The optimal resection for each patient was defined as the smallest resection leading to at least a $90\%$ reduction in seizure propagation. Results The individualized model reproduced the basic aspects of seizure propagation for 9 out of 10 patients when using the resection area as the origin of epidemic spreading, and for 10 out of 10 patients with an alternative definition of the seed region. We found that, for 7 patients, the optimal resection was smaller than the resection area, and for 4 patients we also found that a resection smaller than the resection area could lead to a 100% decrease in propagation. Moreover, for two cases these alternative resections included nodes outside the resection area. Conclusion Epidemic spreading models fitted with patient specific data can capture the fundamental aspects of clinically observed seizure propagation, and can be used to test virtual resections in silico. Combined with optimization algorithms, smaller or alternative resection strategies, that are individually targeted for each patient, can be determined with the ultimate goal to improve surgery outcome.

The role of epidemic spreading in seizure dynamics and epilepsy surgery

“…Here we consider an epidemic spreading model to generate individualized seizure propagation models that are based on the patient-specific MEG connectivity and seizure propagation pathways as derived from invasive EEG recordings. This framework generalizes on previous studies by our group [33,34] by including a recovery mechanism in the spreading model, allowing the return to the healthy (post-ictal) state, so that seizures may remain local (i.e. if the affected regions recover before spreading the ictal state to distant regions) or generalize.…”

“…The emerging behavior of the system under this dynamics is well-characterized in relation to the underlying network structure [51, 55]. In this scenario, the model does not try to mimic the detailed biophysical processes involved in seizure generation and propagation, instead it is used here as an abstraction that includes only the most relevant features of seizure propagation [33, 34, 44, 51]. The model is characterized by two control parameters, the global spreading rate β characterizing the probability of spreading of the infected state, and the recovery rate γ characterizing the recovery probability of each infected node.…”

“…In order to solve this problem, in-depth studies to characterize the dynamical properties of the models, and the interplay between network structure and emergent dynamics, are needed [45–47], often in combination with elaborate modelling optimization frameworks, such as Bayesian inference [43, 44, 48, 49] or deep learning [50]. Another possibility to circumvent these issues is by considering simpler, abstract models that focus only on the behavior of interest: the propagation of ictal activity throughout the brain [33, 34], accounted for by the slow permittivity variable of the epileptor and similar highly dimensional models Sip et al [44]. Conceptually, this process is equivalent to other spreading processes on networks, a problem that has been well-characterized by means of epidemic spreading models [51].…”

“…In previous studies we considered epidemic spreading models as the basis for seizure propagation over the brain, without trying to mimic the complicated biophysical mechanisms involved in the process [33, 34]. Within this framework, we found that epidemic spreading models fitted with patient-specific data could reproduce the stereotypical patterns of seizure propagation on patient-specific brain networks, individually for each patient.…”

“…Within this framework, we found that epidemic spreading models fitted with patient-specific data could reproduce the stereotypical patterns of seizure propagation on patient-specific brain networks, individually for each patient. Moreover, by taking into account this patient-specific connectivity, alternative or smaller resections could be found with the model, which we hypothesized could lead to fewer side effects with the same outcome, in terms of seizure reduction [33, 34].…”