Diagnosing epilepsy is a critical step to outlining treatments. Each patient can have a unique situation where different parts of the brain might be triggering or failing to suppress seizures. Finding which parts of the brain are starting the seizures is an important step towards being able to treat them. Characterizing each patient’s epilepsy condition, however, requires careful monitoring of many events, often requiring patients to stay in the hospital over prolonged perionds, which is both time consuming and expensive. The data from these long and many sessions need to then be carefully annotated and studied by the doctors in order to reach a conclusion on where the siezures are originating from. Reviewing days’ worth of data and the unpredictable ways that seizures can arise is also a recipe for significant errors in diagnoses.

New advances in neuroscience and bioengineering, funded by the National Institutes of Health, are finding new ways to reduce errors and the burden of diagnoses on the doctors, allowing them to more accurately diagnose more patients. These new tools are not based on waiting for seizures and then measuring changes in activity, but rather are based on thinking of the brain as a network and trying to measure when a network is healthy or not. This means that doctors might not have to wait for seizures, reducing monitoring time. The method is based on analyzing electroencephalography (EEG) signals to perform the diagnosis, and compared to traditional approaches, is reported to be up to 3x faster and reduces the number of wrong diagnoses by up to 70%. This exciting new approach is currently undergoing clinical trials in preparation for broader use.