CardioGrid: ECG analysis on demand to detect cardiovascular abnormalities

As the number of cardiovascular related deaths is increasing not only in the western world but also in the developing countries, and there are only limited cardiac diagnosis related resources, it is expected that many healthcare centres will turn to automated diagnosis of ECGs for large number of patients. While systems and tools have been proposed and developed for a handful of cardiovascular diseases, no integrated system exists that can process and analyze very large number of patients on a powerful networked platform to serve healthcare centers online in a cost effective manner. In this paper, we therefore, propose an architecture of a CardioGrid system that allows large number of healthcare centers to submit electrocardiogram (ECGs) of their patients and receive analyzed and annotated reports online. Various functional components of the system have been explained to show how the system would work in real-life. A prototype has been built that is capable of diagnosing numerous cardiovascular diseases, but the proposed architecture also allows the system to continuously learn abnormal patterns from submitted ECGs and annotations from cardiologists. We expect such a system to be useful not only for serious cardiac patients and ageing populations, but also for people that want to get their status checked for wellness monitoring.