Electrocardiogram is a representation of the electric activity of the heart. A medical practitioner can diagnose various heart-related problems by identifying abnormalities in the ECG recording. It takes around 5-10 minutes to take an ECG and the recording is quite lengthy. If the person has infrequent irregularities in heart activity, it is supposed to wear a device called Holter monitor to record the ECG for 24 hrs.
That will generate a lengthier recording. As a result, the diagnosis process entirely depends on the knowledge and experience of the medical practitioner.
Machine learning has been used in this context to assist the human to not only to identify arrhythmias in ECG, but also to integrate with Holter monitor and automated external defibrillators to make intelligent decisions based on the ECG.
This blog points you to a presentation that was delivered as a partial requirement of graduate level Machine Learning course at SUTD. In this presentation, the common deep learning models have been discussed towards ECG arrhythmia classification without considering shallow learning models.
Enjoy!!!
Comments