Epilepsy surgery is the treatment of choice for drug-resistant epilepsy patients, but up to 50% of patients continue to have seizures one year after the resection. In order to aid presurgical planning and predict postsurgical outcome on a patient-by-patient basis, we developed a framework of individualized computational models that combines epidemic spreading with patient-specific connectivity and epileptogeneity maps: the Epidemic Spreading Seizure and Epilepsy Surgery framework (ESSES). ESSES parameters were fitted in a retrospective study (N = 15) to reproduce invasive electroencephalography (iEEG)-recorded seizures. ESSES reproduced the iEEG-recorded seizures, and significantly better so for patients with good (seizure-free, SF) than bad (nonseizure-free, NSF) outcome. We illustrate here the clinical applicability of ESSES with a pseudo-prospective study (N = 34) with a blind setting (to the resection strategy and surgical outcome) that emulated presurgical conditions. By setting the model parameters in the retrospective study, ESSES could be applied also to patients without iEEG data. ESSES could predict the chances of good outcome after any resection by finding patient-specific model-based optimal resection strategies, which we found to be smaller for SF than NSF patients, suggesting an intrinsic difference in the network organization or presurgical evaluation results of NSF patients. The actual surgical plan overlapped more with the model-based optimal resection, and had a larger effect in decreasing modeled seizure propagation, for SF patients than for NSF patients. Overall, ESSES could correctly predict 75% of NSF and 80.8% of SF cases pseudo-prospectively. Our results show that individualised computational models may inform surgical planning by suggesting alternative resections and providing information on the likelihood of a good outcome after a proposed resection. This is the first time that such a model is validated with a fully independent cohort and without the need for iEEG recordings.

There is an increasing awareness of the advantages of multi-modal neuroimaging. Networks obtained from different modalities are usually treated in isolation, which is however contradictory to accumulating evidence that these networks show non-trivial interdependencies. Even networks obtained from a single modality, such as frequency-band specific functional networks measured from magnetoencephalography (MEG) are often treated independently. Here, we discuss how a multilayer network framework allows for integration of multiple networks into a single network description and how graph metrics can be applied to quantify multilayer network organisation for group comparison. We analyse how well-known biases for single layer networks, such as effects of group differences in link density and/or average connectivity, influence multilayer networks, and we compare four schemes that aim to correct for such biases: the minimum spanning tree (MST), effective graph resistance cost minimisation, efficiency cost optimisation (ECO) and a normalisation scheme based on singular value decomposition (SVD). These schemes can be applied to the layers independently or to the multilayer network as a whole. For correction applied to whole multilayer networks, only the SVD showed sufficient bias correction. For correction applied to individual layers, three schemes (ECO, MST, SVD) could correct for biases. By using generative models as well as empirical MEG and functional magnetic resonance imaging (fMRI) data, we further demonstrated that all schemes were sensitive to identify network topology when the original networks were perturbed. In conclusion, uncorrected multilayer network analysis leads to biases. These biases may differ between centres and studies and could consequently lead to unreproducible results in a similar manner as for single layer networks. We therefore recommend using correction schemes prior to multilayer network analysis for group comparisons.