Epilepsy is an archetypal brain network disorder. Despite two decades of research
elucidating network mechanisms of disease and correlating these with outcomes, the clinical
management of children with epilepsy does not readily integrate network concepts. For
example, network measures are not used in presurgical evaluation to guide decision making
or surgical management plans.
The aim of this thesis was to investigate novel network frameworks from the perspective of
a clinician, with the explicit aim of finding measures that may be clinically useful and
translatable to directly benefit patient care. We examined networks at three different scales,
namely macro (whole brain diffusion MRI), meso (subnetworks from SEEG recordings) and
micro (single unit networks) scales, consistently finding network abnormalities in children
being evaluated for or undergoing epilepsy surgery. This work also provides a path to clinical
translation, using frameworks such as IDEAL to robustly assess the impact of these new
technologies on management and outcomes.
The thesis sets up a platform from which promising computational technology, that utilises
brain network analyses, can be readily translated to benefit patient care.