Number of the records: 1

Classification of ECG using ensemble of residual CNNs with or without attention mechanism

SYSNO ASEP	0557480
Document Type	J - Journal Article
R&D Document Type	Journal Article
Subsidiary J	Článek ve WOS
Title	Classification of ECG using ensemble of residual CNNs with or without attention mechanism
Author(s)	Nejedlý, Petr (UPT-D)_{RID, SAI} Ivora, Adam (UPT-D) Viščor, Ivo (UPT-D)_{RID, ORCID, SAI} Koščová, Zuzana (UPT-D) Smíšek, Radovan (UPT-D)_{RID, ORCID, SAI} Jurák, Pavel (UPT-D)_{RID, ORCID, SAI} Plešinger, Filip (UPT-D)_{RID, ORCID, SAI}
Number of authors	7
Article number	044001
Source Title	Physiological Measurement. - : Institute of Physics Publishing - ISSN 0967-3334 Roč. 43, č. 4 (2022)
Number of pages	12 s.
Publication form	Print - P
Language	eng - English
Country	GB - United Kingdom
Keywords	ECG ; classification ; deep learning ; PhysioNet challenge 2021 ; attention mechanism
Subject RIV	FS - Medical Facilities ; Equipment
OECD category	Medical engineering
R&D Projects	FW01010305 GA TA ČR - Technology Agency of the Czech Republic (TA ČR)
Method of publishing	Limited access
Institutional support	UPT-D - RVO:68081731
UT WOS	000790542900001
EID SCOPUS	85129997154
DOI	10.1088/1361-6579/ac647c
Annotation	Objective. This paper introduces a winning solution (team ISIBrno-AIMT) to the official round of PhysioNet Challenge 2021. The main goal of the challenge was a classification of ECG recordings into 26 multi-label pathological classes with a variable number of leads (e.g. 12, 6, 4, 3, 2). The main objective of this study is to verify whether the multi-head-attention mechanism influences the model performance. Approach. We introduced an ECG classification method based on the ResNet architecture with a multi-head attention mechanism for the official round of the challenge. However, empirical findings collected during model development suggested that the multi-head attention layer might not significantly impact the final classification performance. For this reason, during the follow-up round, we removed a multi-head attention layer to test the influence on model performance. Like the official round, the model is optimized using a mixture of loss functions, i.e. binary cross-entropy, custom challenge score loss function, and custom sparsity loss function. Probability thresholds for each classification class are estimated using the evolutionary optimization method. The final architecture consists of three submodels forming a majority voting classification ensemble. Main results. The modified model without the multi-head attention layer increased the overall challenge score to 0.59 compared to the 0.58 from the official round. Significance. Our findings from the follow-up submission support the fact that the multi-head attention layer in the proposed architecture does not significantly affect the classification performance.
Workplace	Institute of Scientific Instruments
Contact	Martina Šillerová, sillerova@ISIBrno.Cz, Tel.: 541 514 178
Year of Publishing	2023
Electronic address	https://iopscience.iop.org/article/10.1088/1361-6579/ac647c

Number of the records: 1