A Generative Modeling Approach to Limited Channel ECG Classification

Rajan, Deepta; Thiagarajan, Jayaraman J.

doi:10.1109/embc.2018.8512757

Cited by 34 publications

(22 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several cardiac diseases including atrial fibrillation, ventricular flutter, tachycardias and left/right bundle branch blocks (BBB) manifest as anomalous deviations in ECG channel configurations, where each lead provides a unique perspective on the electrical activity of the heart. For example, leads II, III, and aVF are used to detect inferior myocardial infarction, while leads V1 and V6 are used for bundle branch block 8 . Despite that, in telemetry and other ambulatory settings only a subset of channels are accessible, making the task of abnormality detection more challenging 9 .…”

Section: Ecg-based Arrhythmia Classification a Crucial Step Towards mentioning

confidence: 99%

“…Detecting myocardial infarction (MI) is one of the most crucial problems within the computerized ECG interpretation community. There have been several solutions proposed to solve this task using both 12 ECG channels as in 29,30 and 19 , as well as a limited subset of channels as shown in 8,18 . The standard 12-lead ECG depicts evidence of ischemic heart diseases that predominantly occur due to the narrowing of blood vessels caused by atherosclerosis.…”

Section: Ecg-based Myocardial Infarction Detectionmentioning

confidence: 99%

See 1 more Smart Citation

DDxNet: a deep learning model for automatic interpretation of electronic health records, electrocardiograms and electroencephalograms

Thiagarajan

Rajan

Katoch

et al. 2020

Sci Rep

Self Cite

View full text Add to dashboard Cite

Effective patient care mandates rapid, yet accurate, diagnosis. With the abundance of non-invasive diagnostic measurements and electronic health records (EHR), manual interpretation for differential diagnosis has become time-consuming and challenging. This has led to wide-spread adoption of AI-powered tools, in pursuit of improving accuracy and efficiency of this process. While the unique challenges presented by each modality and clinical task demand customized tools, the cumbersome process of making problem-specific choices has triggered the critical need for a generic solution to enable rapid development of models in practice. In this spirit, we develop DDxNet, a deep architecture for time-varying clinical data, which we demonstrate to be well-suited for diagnostic tasks involving different modalities (ECG/EEG/EHR), required level of characterization (abnormality detection/phenotyping) and data fidelity (single-lead ECG/22-channel EEG). Using multiple benchmark problems, we show that DDxNet produces high-fidelity predictive models, and sometimes even provides significant performance gains over problem-specific solutions.

show abstract

Section: Ecg-based Arrhythmia Classification a Crucial Step Towards mentioning

confidence: 99%

Section: Ecg-based Myocardial Infarction Detectionmentioning

confidence: 99%

DDxNet: a deep learning model for automatic interpretation of electronic health records, electrocardiograms and electroencephalograms

Thiagarajan

Rajan

Katoch

et al. 2020

Sci Rep

Self Cite

View full text Add to dashboard Cite

show abstract

“…A series of comparative experiments were conducted to assess conventional models in the classification of multichannel process signals. This included a multi-channel convolutional network (MC-DCNN) [40], an LSTM + RF network combining LSTM and random forest [41], and a GRU-RNN network [3]. These three deep neural network models, combined with the methodology proposed in this study, comprised the four algorithms used for comparative disease classification.…”

Section: ) Comparative Analysismentioning

confidence: 99%

“…W ITH the rapid development and application of intelligent sensor and internet of things (IOT) technology, multi-channel signal classification has become increasingly important in a variety of fields [1]- [3]. Nonlinear timevarying signals are a type of multi-component signal whose frequency and amplitude change with time.…”

Section: Introductionmentioning

confidence: 99%

A CNN-RBPNN Model With Feature Knowledge Embedding and its Application to Time-Varying Signal Classification

Lü¹,

Wang²,

Xu³

et al. 2020

IEEE Access

View full text Add to dashboard Cite

A novel technique, combining the feature extraction mechanisms of a convolutional neural network (CNN) with the classification method of a radial basis probability neural network (RBPNN), is proposed for small sample set modeling and feature knowledge embedding in multi-channel time-varying signal classification. This CNN-RBPNN consists of a signal input layer, signal feature parallel extraction and integration units, and an RBPNN classifier. Each channel signal in a feature extraction unit corresponds to a 1D CNN. The extracted features are represented as feature vectors, and these vectors constitute a comprehensive feature matrix. The RBPNN classifier was designed using signal feature embedding mechanism based on radial basis kernels and the property of combining pattern subclasses into pattern classes to form complex class boundaries. A dynamic clustering algorithm was used to divide each pattern class sample into several subclasses. Typical signal samples in each pattern subclass were designated as kernel centers, in order to achieve signal categories features embedding. This process was also used to determine the number of nodes in the RBPN layer. The RBPN layer outputs were selectively summed in the pattern layer according to kernel center category, which can generate irregular class boundaries, reducing the overlap among different pattern class boundary. The proposed CNN-RBPNN replaces the fullconnection layer and classifier unit of conventional CNN with RBPNN, which can extract and represent signals distribution features and structural properties, implement structural and data constraints. This can reduce the structural risks of small sample set modeling. In this study, the properties of CNN-RBPNN are analyzed and an integrated learning algorithm is proposed. An experiment was conducted using 12-lead ECG signals in a seven-classification in the case of small sample set. Results demonstrated that, the correct recognition rate is 5.7% higher than other methods in the experiment, the performance evaluation index also showed significant improvement. INDEX TERMS Time-varying signal classification, CNN feature extraction, radial basis probability network, prior feature knowledge embedding, small sample set modeling

show abstract

“…In 2017, Rajpurkar et al used a 34-layer CNN model to diagnose 12 diseases based on 30,000 ECG signal samples with single-lead and 30s duration in Zio Patch Monitor Dataset, and achieved an accuracy of 80.9% [33]. Schwab [37]. Comprehensive analysis, the above researches have achieved high recognition accuracies.…”

Section: E Comparison and Analysismentioning

confidence: 99%

A Probabilistic Process Neural Network and Its Application in ECG Classification

Feng

Liang

et al. 2019

IEEE Access

View full text Add to dashboard Cite

A novel probabilistic process neural network (PPNN) model is proposed for the multi-channel time-varying signal classification problems with ambiguity and randomness distribution characteristics. This model was constructed from an input time-varying signal layer, a probabilistic process neuron (PPN) hidden layer, a pattern layer, and a Softmax classifier. The number of nodes in the input layer is the same as the number of time-varying signal input channels, which can realize the overall input of the time-varying signal. The PPN hidden layer performs the spatio-temporal aggregation operations for time-varying input signals and probabilistic outputs. The connection weight functions from the input layer to the hidden layer are represented by the typical samples or cluster center functions in different pattern subsets of the sample set, which the prior knowledge of signal categories is implicitly expressed by the morphological distribution features and combinational relations. A pattern layer selectively summed to the output of the PPN hidden layer using the categorical attributes of the connection weight function vector. And the Softmax classifier implements the probabilistic classification of time-varying signals. PPNN has the advantages of fewer model parameters, suitable for modeling small samples and integrating prior knowledge of signal categories. This paper develops the specific learning algorithms which synthesize dynamic time warping, C-means clustering, and BP algorithm. The 12-lead electrocardiogram (ECG) signals for heart disease diagnosis were used for classification testing. The experimental results from 12-lead ECG signal across ten types of disease classification verify the effectiveness of the model and the proposed algorithm. INDEX TERMS Dynamic signal classification, Bayesian decision rules, probabilistic process neural network, training algorithm, ECG signal classification.

show abstract

A Generative Modeling Approach to Limited Channel ECG Classification

Cited by 34 publications

References 13 publications

DDxNet: a deep learning model for automatic interpretation of electronic health records, electrocardiograms and electroencephalograms

DDxNet: a deep learning model for automatic interpretation of electronic health records, electrocardiograms and electroencephalograms

A CNN-RBPNN Model With Feature Knowledge Embedding and its Application to Time-Varying Signal Classification

A Probabilistic Process Neural Network and Its Application in ECG Classification

Contact Info

Product

Resources

About