EEG data handling using AWS cloud computing for automatic spike detection in epilepsy patients

Abstract

In today’s era, healthcare and medical applications encountering rapid growth in data. The key challenges involved are effectively managing and handling such a high volume of diverse medical data. Epilepsy is considered as one of the most serious brain diseases around the world. Electrophysiological data, such as EEG can be called ‘big data’ as they might have more than 50 multi-channel signals from each patient, generating more than 5 GB data. An adequate approach to store and analyse signal data is required to meet the enormous volume of EEG data. To develop a system of a very fast web-based EEG browser for the analysis of the huge amount of EEG data, a cloud-based platform is introduced to store and automate EEG data interpretation with the help of machine learning techniques on a web-based EEG browser. In this work, we explore the feasibility of AWS-cloud based approach for handling large EEG data for automatic spike detection in epilepsy patients. The aim is to design a system which will allow neurologists to upload and download EEG data from any region for analysis purpose; in the future the data will be analysed by machine learning algorithms. The approach is to store large EEG data in one place for an example, in the AWS cloud so that many users can upload and have a quick access to the data from different regions of the world. AWS cloud computing provides a simple way to access the storage, databases and offers a set of application services.