•  
  •  
 

Abstract

Long short-term memory networks can effectively process complex temporal patterns in electrocardiogram data. These sequential models excel at classifying heart disease from the rich signals captured by electrocardiograms. However, traditional algorithms struggle with the intricate waveforms encoded in each heartbeat. Deeper architectures such as LSTM are better equipped to untangle the subtle variations between healthy sinus rhythms and lethal arrhythmias.

In this study, an LSTM model was developed to diagnose disease from the PTB dataset. The network was trained using a fusion of deep learning schemes for sequential data. The model underwent several evaluations, from a confusion matrix mapping predictions to ground truths, to numerical metrics quantifying accuracy, F1 score and memory retention over time. Qualitative insights also revealed how the system differentiated classes to improve understanding.

The results showed the network could diagnose various heart conditions with up to 97.9% accuracy and was better at distinguishing patient classes than older techniques. In addition, the network excelled at detecting subtle but critical patterns in electrocardiograms, enhancing medical diagnosis for improved care.

Share

COinS