Cardiac arryhthmia identification system

Mónica Yolanda Moreno Revelo and Sandra Carolina Patascoy Botina, Universidad de Nariño, San Juan de Pasto-Colombia 2017

Cardiac arrhythmias are heart condition alterations due mainly to the change of heart rate, generated when heart’s electric conduction system works improperly.The AAMI (Association for the advanced medical of instrumentation) recommend to study a set of heartbeats such as: Normal beats (N), Supraventricular ectopic beat (S), Ventricular ectopic beat (V), Fusion beat (F), and unknown beat class (Q).

Figure 1. Types Heartbeats recommended

The automatic diagnostic consists of classifying heartbeats into groups holding heartbeats belonging to the same class (any pathology or normal). In terms of machine learning classification, it can be performed by techniques from either supervised or unsupervised analysis; being the latter the most recommended. Since they allows for reducing the amount of heartbeats that a cardiologist should revise, unsupervised techniques result advisable for this classification problem. 

Figure 2. Grouped Heartbeats

Graphic Interface

Proposed interface to visualize the hearbeat clustering process.


Cardiac arryhthmia identification system based on non-supervised machine learning methods.

International Conference on Information Systems and Computer Science (INSISCOS)

see full paper

Design of an automatic identification of arrhythmias system using Unsupervised techniques

International society for computational biology (ISCB)


Tutorial Video

About us

 Mónica Moreno Sandra Patascoy
 Electronic Engineer Electronic Engineer