Recent advances in the field of Ambulatory Electrocardiography (A-ECG) have necessitated the creation of a number of Cardiac Arrhythmia Databases (CADs) for evaluating A-ECG analysis systems performance. Although the two most widely recognized CADs were developed independently by the Massachusetts Institute of Technology (MIT) and the American Heart Association (AHA), both suffer from limitations either in size (MIT) or numbers of arrhythmia classes (AHA). The proposed research is to ascertain the feasibility of designing and developing a CAD which will overcome these limitations, and provide an extensive Arrhythmia Library System (ALS) accessible for multiple clinical, academic, and industrial applications. Most of the technical, clinical, and administrative protocols will be designed around a hybrid combination of those utilized by MIT and the AHA. In addition, new protocols will be created which will allow for increases in the classes of arrhythmias, and the recording duration of digitized annotated A-ECGs. A standardized protocol for preliminary analysis, classification, and reporting of A-ECG data from six national analysis centers will be designed to enable the development of an Arrhythmia Screening Database (ASD) representing approximately 3,000 recording per month. From this monthly pool, over 30 specific arrhythmia classes and associated demographics will be identified for ALS inclusion.
