Show simple item record

dc.contributor.authorSigurthorsdottir, Halla
dc.contributor.authorVan Zaen, Jérôme
dc.contributor.authorDelgado-Gonzalo, Ricard
dc.contributor.authorLemay, Mathieu
dc.date.accessioned2022-02-14T17:08:20Z
dc.date.available2022-02-14T17:08:20Z
dc.date.issued2020-12-30
dc.identifier.citation2020 Computing in Cardiology Conference, Rimini (Italy)
dc.identifier.urihttps://yoda.csem.ch/handle/20.500.12839/993
dc.description.abstractWe 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 ?nal 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 ?lters 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 24th out of 41 in the of?cial ranking.
dc.titleECG Classification With a Convolutional Recurrent Neural Network
dc.typeProceedings Article
dc.type.csemdivisionsDiv-E
dc.type.csemresearchareasDigital Health
dc.identifier.urlhttp://www.cinc.org/archives/2020/pdf/CinC2020-198.pdf
dc.identifier.doi10.22489/CinC.2020.198


Files in this item

FilesSizeFormatView

There are no files associated with this item.

This item appears in the following Collection(s)

  • Research Publications
    The “Research Publications” collection provides bibliographic information for scientific papers including conference proceedings and presentations.

Show simple item record