Deep Learning for Cardiologist-level Myocardial Infarction Detection in Electrocardiograms
Problem: To detect risk of heart disease from ECG signals.
Importance: To provide a correct diagnosis so that proper health care can be given to patients.
Area: Deep learning – scientific machine learning
o Concerned about design, training, and use of ML algorithm in an optimal matter towards a certain problem.
Deep learning model benefits:
o Uses multiple GPUs to construct complicated NN that can be trained using TB-sized datasets, which are robust against noise.
o After training is complete, it takes little computational power to conduct statistical inference.
Question that need to be answered:
o What AI architecture and datasets can provide evidence of symptoms, ECG changes, imaging evidence that shows loss of viable myocardium and wall motion abnormality?
Purpose of the article:
o To provide a detailed analysis on the contribution of each ECG lead in identifying heart disease in the model. This is because the selection of data in previous studies regarding heart disease made data selection arbitrary in identifying heart conditions.
o To show the use of using multiple data channels of information to enhance prediction accuracy in deep learning, i.e. processing the top three ECG leads simultaneously in the neural network.
o Show that feature engineering is not necessary in the training, validation, or testing process for ECG data in neural networks.
Database used: PTB database
1. CNN Network
a. Used both noisy and denoised ECGs without feature engineering.
2. Used artificial neural network, probabilistic neural network, KNN, multi-layer perceptron, and Naïve Bayes Classification
a. Extracted two features: T-wave integral and total integral to identify heart disease.
3. Developed two different ANN: RBF and MLP
4. Supervised learning techniques have limited success to the problem and used multiple instance learning.
a. Demonstrate proposed algorithm LTMIL surpasses supervised approaches.
5. Create new feature by approximating ECG signal using a 20th order polynomial, which achieved 94.4% accuracy.
6. Stationary wavelet transforms to decompose ECG into sub-bands.
a. SVM and KNN used to classify.
b. Features used: sample entropy, normalized sub-bands, log energy entropy, median slope from sub-bands.
7. Transfer learning – used deep CNN model for arrhythmia and developed it to detect heart disease.
8. Simple adaptive threshold (SAT)
a. Multiresolution approach, adaptive thresholding used to extract features: depth of Q peak and elevation in ST segment
9. Subject-oriented approach using CNN to take in leads II, III, AVF
10. Model ECG using 2nd order ODE and feed the best-fitting coefficients of the ECG signal into a SVM.
11. Multi-channel CNN (16 layers) with long-short memory units
12. Deep CNN that takes 3 seconds at a time of lead II as input
Most use feature extraction/selection from the raw ECG data before training.
o Problem with feature selection is that it is not practical for large volumes of data.
Other papers that do not use feature selection arbitrarily picks ECG leads for classification and does not provide rationale.
1. Quantification of accuracies for single channels with 20-fold cross-validation, resulting highest individual accuracies: v5, v6, vx, vz, and ii
2. Quantification of accuracies for pairs of top 5 highest individual channels with 20-fold cross-validation, resulting highest pairs accuracies to fed into a the neural network: lead v6 and lead vz
3. Use 100-fold cross validation on v6 and vz pair of channels, then compare outliers based on top 20, top 50 and total 100 performing models, finding that standard deviation is non-trivial and there are few models performed very poorly.
4. Discussing 2 factors effecting model performance evaluation:
1） Random train-val-test split might have effects of the performance of the model, but it can be improved by access with a larger data set and further discussion
2） random initialization of the weights of neural network shows little effects on the performance of the model performance evaluation, because of showing a high average results with a fixed train-val-test split
5. Comparing with other models, the model in this article has the highest accuracy, specificity, and precision
6. Further using 290 fold patient-wise split, resulting the same highest accuracy of the pair v6 and vz as record-wise split
1） Discuss patient-wise split might result lower accuracy evaluation, however, it still maintain an average of 97.83%
3.1: Data curation
Dataset: 549 ECG records total
290 unique patients
Each ECG record has a mean length of over 100s
3.2: ANN model
ConvNetQuake model + 1D batch normalization + Label-smoothing
- Input layer: 10-second long ECG signal
- Hidden layers: 8 * (1D convolution layer, Activation function: RELU, 1D batch normalization layer)
- Output layer: 1280 dimensions -> 1 dimension, Activation function: Sigmoid
Batch size = 10
Learning rate = 10^-4
Optimizer = ADAM