ECG Abnormalities Recognition Using Convolutional Network With Global Skip Connections and Custom Loss Function

Vičar, Tomáš; Hejc, Jakub; Novotná, Petra; Ronzhina, Marina; Janoušek, Oto

doi:10.22489/cinc.2020.189

Cited by 14 publications

(8 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The first loss function uses binary cross-entropy to train the weights of the model. The second loss function utilizes a challenge-specific loss function proposed by Vicar et al (2020) during last year's challenge. In general, these metrics directly optimize challenge scores by computing a differential approximation of proposed challenge metrics.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

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

Nejedlý¹,

Ivora²,

Viščor³

et al. 2022

Physiol. Meas.

View full text Add to dashboard Cite

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 variable number of leads (e.g., 12,6,4,3,2). 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. 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. 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.

show abstract

Section: Methodsmentioning

confidence: 99%

“…Also, we considered adding a custom challenge loss function proposed by Vicar et al (2020), where the continuous equivalent of binary OR operator was used to design a differentiable approximation of challenge metric. But we believed that further modification of this loss function would be beneficial, considering binary data output.…”

Section: Introductionmentioning

confidence: 99%

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

Nejedlý¹,

Ivora²,

Viščor³

et al. 2022

Physiol. Meas.

View full text Add to dashboard Cite

show abstract

“…A temporal convolutional neural network (1D-CNN) handling ECG channels as a separate features (Vicar et al, 2020(Vicar et al, , 2021 was employed to learn a non-linear mapping h : X → H, where h is a feature extractor and H…”

Section: Embeddings Extractionmentioning

confidence: 99%

Learning Discriminative Representations of Superimposed P waves With Weakly-Supervised Temporal Contrastive Learning

Hejc,

Redina,

Pospisil

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

Electrocardiography (ECG) wave morphology and timing provide critical diagnostic information for diagnosing arrhythmias and conduction abnormalities, allowing risk stratification for various cardiac diseases. However, accurate extraction of these features becomes challenging in the presence of superimposed waves from distinct cardiac chambers, a common occurrence during pathological rhythms. This work proposes a novel Surrogate-boosted Temporal Contrastive Representation Learning (S-TCRL) framework to address this challenge. S-TCRL leverages weak labels, readily obtainable from invasive catheter examinations, to extract latent representations of superimposed P waves. We reformulate the problem from object-wise to sample-wise incomplete information by employing surrogate labels. A 1D fully-convolutional feature pyramid network (FPN) extracts multi-scale features from ECG signals. These features are segmented into equal-sized temporal regions, whose labels are inferred from individual samples using a multiple-instance learning (MIL) paradigm. Non-sequential embeddings are generated to facilitate alignment-free cosine similarity estimation. A temperature-scaled cross-entropy loss function minimizes the distance between embeddings of similar regions (likely containing P waves) while maximizing the distance between dissimilar ones. The framework's efficacy is evaluated on a custom ECG dataset comprising 3265 short-term recordings from 708 individuals undergoing catheter ablation. S-TCRL achieves significant improvement in the downstream P wave segmentation task compared to two baseline MIL methods. The average recall and precision for both P wave boundaries reach 70.0% and 80.0%, respectively, exceeding the baselines' 63.5% and 67.5%. These results demonstrate the potential of S-TCRL for embedding representation of superimposed P waves and its potential generalizability to other tasks like arrhythmia classification.

show abstract

“…Augmentation techniques generate new training samples by adding some perturbation in data resulting in improved robustness of the model. First, some manipulations can be applied on initial data, such as random scaling, flipping, shifting, and noising ECG, to achieve accurate detection of multiple arrhythmias ( Vicar et al, 2020 ; Nonaka and Seita, 2021 ; Do et al, 2022 ). The same application can profit from using the synthetic samples generated from the training ones using intuitive adaptive synthetic data sampling (ADASYN, Virgeniya and Ramaraj, 2021 ) or synthetic minority oversampling technique (SMOTE, Ketu and Mishra, 2021 ).…”

Section: Ecg Analysismentioning

confidence: 99%

Golden Standard or Obsolete Method? Review of ECG Applications in Clinical and Experimental Context

et al. 2022

Self Cite

View full text Add to dashboard Cite

Cardiovascular system and its functions under both physiological and pathophysiological conditions have been studied for centuries. One of the most important steps in the cardiovascular research was the possibility to record cardiac electrical activity. Since then, numerous modifications and improvements have been introduced; however, an electrocardiogram still represents a golden standard in this field. This paper overviews possibilities of ECG recordings in research and clinical practice, deals with advantages and disadvantages of various approaches, and summarizes possibilities of advanced data analysis. Special emphasis is given to state-of-the-art deep learning techniques intensely expanded in a wide range of clinical applications and offering promising prospects in experimental branches. Since, according to the World Health Organization, cardiovascular diseases are the main cause of death worldwide, studying electrical activity of the heart is still of high importance for both experimental and clinical cardiology.

show abstract

ECG Abnormalities Recognition Using Convolutional Network With Global Skip Connections and Custom Loss Function

Cited by 14 publications

References 9 publications

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

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

Learning Discriminative Representations of Superimposed P waves With Weakly-Supervised Temporal Contrastive Learning

Golden Standard or Obsolete Method? Review of ECG Applications in Clinical and Experimental Context

Contact Info

Product

Resources

About