ECG Classification with a Convolutional Recurrent Neural Network

TL;DR

This paper presents a convolutional recurrent neural network for classifying 12-lead ECG signals, combining convolutional feature extraction with recurrent aggregation, optimized hyperparameters, and test time augmentation, achieving competitive challenge scores.

Contribution

The paper introduces a novel CNN-RNN architecture with attention for ECG classification, optimizing hyperparameters and applying test time augmentation for improved performance.

Findings

Achieved a challenge score of 0.511 on validation set.

Ranked 23rd out of 41 in the challenge.

Optimized hyperparameters significantly impacted model performance.

Abstract

We developed a convolutional recurrent neural network to classify 12-lead ECG signals for the challenge of PhysioNet/ Computing in Cardiology 2020 as team Pink Irish Hat. The model combines convolutional and recurrent layers, takes sliding windows of ECG signals as input and yields the probability of each class as output. The convolutional part extracts features from each sliding window. The bi-directional gated recurrent unit (GRU) layer and an attention layer aggregate these features from all windows into a single feature vector. Finally, a dense layer outputs class probabilities. The final decision is made using test time augmentation (TTA) and an optimized decision threshold. Several hyperparameters of our architecture were optimized, the most important of which turned out to be the choice of optimizer and the number of filters per convolutional layer. Our network achieved a challenge score of 0.511 on the hidden validation set and 0.167 on the full hidden test set, ranking us 23rd out of 41 in the official ranking.