Development of a Natural Language Processing pipeline to support the differential diagnosis of epilepsy based on the descriptions of the semiology of the seizures
According to World Health Organization (WHO), epilepsy is one of the most common serious neurological disorder. It affects 50 million individuals worldwide and up to 10% of people have one seizure during their lifetime. So, based on the information available in the Electronic Health Records (EHRs) from the CRIS database the main aim of this project is to build a support tool for epilepsy diagnosis considering seizures as symptoms. Once a document is classified with a good degree of reliability, the description of the specific signs of each seizure will be used to distinguish between different epilepsy classes.