CARDIOBLINK - A Secure Web-Based Platform for ECG Analysis and Cardiac Disease Detection

This paper introduces the CARDIOLINK — a safe online foundation for convenient ECG analysis and heart diseases diagnosis. Security of access is the prime concern in this system hence only administrators are allowed to login. When an admin logs in, they are directed to a page which permits them to upload files that contain ECG data. The formats acceptable on this site include three; .dat file (data file), .hea file (header file) and .atr file (attribute file). Once the upload is done, it verifies if the uploaded files have been corrupted or not before giving feedback. Users can make this analysis start by pressing "Analyze" button after uploading the files. CARDIOLINK then displays two sets of patient’s ECG waveforms captured at different time intervals for comparison purposes to analyze them comparatively with each other. Upon reviewing both electrocardiograms, potential heart diseases suffered by a patient are indicated on results page viewable by user. Also, platform provides other details such as common symptoms; names of recommended drugs; factors of risks attached to identified disease and preventive measures. According to the research, done up to now, random forest algorithm is used beneath for the sake of disease foretelling because it robustly survives noise and gives high accuracy rate on large dataset like those produced by electrocardiogram machine.