Automatic method to delineate or categorize an electrocardiogram

1. A computerized method for detecting abnormalities in an electrocardiogram (ECG) signal obtained from a patient, the computerized method comprising:

receiving the ECG signal sampled at a plurality of time points;

generating at least one matrix representative of at least a portion of the ECG signal;

computing delineation scores of the ECG signal at each time point of the plurality of time points of the at least a portion of the ECG signal, the delineation scores corresponding to multiple wave types associated with the ECG signal at each time point using the at least one matrix;

computing anomaly scores to detect anomalies associated with the ECG signal using the at least one matrix and the delineation scores;

determining that an anomaly is present based on the anomaly scores; and

generating information to indicate the presence of the anomaly associated with the ECG signal based on determining that the anomaly is present.

2. The computerized method of claim 1 , further comprising denoising and removing a baseline of the ECG signal.

3. The computerized method of claim 1 , wherein determining that the anomaly is present comprises determining that atrial fibrillation is present.

4. The computerized method of claim 1 , further comprising causing display of the information to indicate the presence of the anomaly.

5. The computerized method of claim 4 , wherein causing the display further comprises causing display of a representation of the ECG signal along with the information to indicate the presence of the anomaly.

6. The computerized method of claim 1 , wherein computing the anomaly scores comprises applying the at least one matrix to at least one trained neural network to compute the anomaly scores.

7. The computerized method of claim 6 , further comprising training at least one neural network with a dataset of pre-characterized ECG signals to generate the at least one trained neural network.

8. The computerized method of claim 6 , wherein the at least one trained neural network comprises a first neural network for classification.

9. The computerized method of claim 6 , wherein the at least one trained neural network further comprises a second neural network for delineation.

10. The computerized method of claim 1 , further comprising:

determining that at least one additional anomaly is present based on the anomaly scores; and

generating information to indicate the presence of the at least one additional anomaly associated with the ECG signal based on determining that the at least one additional anomaly is present.

11. The computerized method of claim 1 , wherein generating the information to indicate the presence of the anomaly comprises assigning an anomaly label to the ECG signal for the anomaly.

12. The computerized method of claim 1 , wherein determining that the anomaly is present based on the anomaly scores comprises comparing the anomaly scores to at least one threshold value.

13. The computerized method of claim 1 , wherein generating the at least one matrix comprises generating the at least one matrix of size m×n where “m” is a number of leads of a cardiac sensor used to obtain the ECG signal and “n” is a number of time points of the plurality of time points.

14. The computerized method of claim 1 , wherein generating the at least one matrix comprises generating the at least one matrix of size p×n where “p” is a number of cardiac wave types and “n” is a number of time points of the plurality of time points.

15. The computerized method of claim 1 , further comprising expressing a plurality of detected anomalies as a vector of size q, with q being a number of anomalies to identify.

16. The computerized method of claim 1 , further comprising determining whether at least one type of cardiac wave is present at each time point of the plurality of time points.

17. The computerized method of claim 1 , further comprising determining that the at least a portion of the ECG signal is normal based on the anomaly scores.

18. A system for detecting abnormalities in an electrocardiogram (ECG) signal obtained from a patient, the system comprising at least one server and at least one processor configured to execute instructions to:

receive the ECG signal sampled at a plurality of time points;

generate at least one matrix representative of at least a portion of the ECG signal;

compute delineation scores of the ECG signal at each time point of the plurality of time points of the at least a portion of the ECG signal, the delineation scores corresponding to multiple wave types associated with the ECG signal at each time point using the at least one matrix;

compute anomaly scores to detect anomalies associated with the ECG signal using the at least one matrix and the delineation scores;

determine that an anomaly is present based on the anomaly scores; and

generate information to indicate the presence of the anomaly associated with the ECG signal based on the determination that the anomaly is present.

19. The system of claim 18 , wherein to determine that the anomaly is present comprises determine that atrial fibrillation is present.

20. The system of claim 18 , wherein the at least one processor is configured to execute instructions to cause display of the information to indicate the presence of the anomaly.

21. The system of claim 20 , wherein the at least one processor is configured to execute instructions to cause display of a representation of the ECG signal along with the information to indicate the presence of the anomaly.

22. The system of claim 18 , wherein to compute the anomaly scores comprises applying the at least one matrix to at least one trained neural network to compute the anomaly scores.

23. The system of claim 22 , wherein the at least one processor is configured to execute instructions to train at least one neural network with a dataset of pre-characterized ECG signals to generate the at least one trained neural network.

24. The system of claim 22 , wherein the at least one trained neural network comprises a first neural network for classification.

25. The system of claim 24 , wherein the at least one trained neural network further comprises a second neural network for delineation.

26. The system of claim 18 , wherein the at least one processor is configured to execute instructions to:

determine that at least one additional anomaly is present based on the anomaly scores; and

generate information to indicate the presence of the at least one additional anomaly associated with the ECG signal based on the determination that the at least one additional anomaly is present.

27. The system of claim 18 , wherein the at least one processor is configured to execute instructions to determine whether at least one type of cardiac wave is present at each time point of the plurality of time points.

28. The system of claim 18 , wherein the at least one processor is configured to execute instructions to determine that the at least a portion of the ECG signal is normal based on the anomaly scores.

29. A programmed routine for use with a computerized system for detecting abnormalities in an electrocardiogram (ECG) signal obtained from a patient, the programmed routine comprising instructions that when executed:

receive the ECG signal sampled at a plurality of time points;

generate at least one matrix representative of at least a portion of the ECG signal;

compute delineation scores of the ECG signal at each time point of the plurality of time points of the at least a portion of the ECG signal, the delineation scores corresponding to multiple wave types associated with the ECG signal at each time point using the at least one matrix;

compute anomaly scores to detect anomalies associated with the ECG signal using the at least one matrix and the delineation scores;

determine that an anomaly is present based on the anomaly scores; and

generate information to indicate the presence of the anomaly associated with the ECG signal based on the determination that the anomaly is present.

30. The programmed routine of claim 29 , wherein to determine that the anomaly is present comprises determine that atrial fibrillation is present.