A deep-learning classifier for cardiac arrhythmias
This work addresses cardiac arrhythmia classification for healthcare applications, but it is incremental as it improves on existing methods with a more efficient network design.
The authors tackled the problem of classifying heart beats into 13 classes, including arrhythmias, by developing a neural network that localizes QRS peaks and uses convolutional layers scaled to the problem's characteristic size. Their method achieved better results with a smaller network than previous approaches, making it competitive for IoT deployment.
We report on a method that classifies heart beats according to a set of 13 classes, including cardiac arrhythmias. The method localises the QRS peak complex to define each heart beat and uses a neural network to infer the patterns characteristic of each heart beat class. The best performing neural network contains six one-dimensional convolutional layers and four dense layers, with the kernel sizes being multiples of the characteristic scale of the problem, thus resulting a computationally fast and physically motivated neural network. For the same number of heart beat classes, our method yields better results with a considerably smaller neural network than previously published methods, which renders our method competitive for deployment in an internet-of-things solution.