Prediction and prevention of postoperative atrial fibrillation in cardiac surgery patients
Systems and methods are provided for predicting the onset of postoperative atrial fibrillation (AF) from electrocardiogram (ECG) data representing a patient. A signal processing component determines parameters representing the activity of the heart of the patient from the ECG data. A feature extraction component calculates a plurality of features useful in predicting postoperative AF from the determined parameters. A classification component determines an AF index for the patient from the calculated plurality of features. The AF index represents the likelihood that the patient will experience AF.
1. A system for predicting the onset of atrial fibrillation (AF) from an electrocardiogram (ECG) signal representing a patient, comprising:
a signal processing component that determines parameters representing the activity of the heart of the patient from the ECG data, the signal processing component comprising a plurality of R-wave detectors that detect R-waves with respective channels of the ECG signal and a multichannel comparison component that confirms that each R-wave is detected in a threshold number of channels;
a feature extraction component that calculates a plurality of features from the determined parameters, at least one feature being calculated from the detected R-waves; and
a classification component that determines an AF index for the patient, representing the likelihood that the patient will experience AF, from the calculated plurality of features.
2. The system of claim 1 , each R-wave detector comprising a wavelet-based filter that removes noise and minimizes false identification of R-waves within an associated channel of the ECG signal.
3. The system of claim 1 , the classification component comprising a hybrid neuro-fuzzy classification system.
4. The system of claim 1 , the feature extractor calculating at least one feature representing premature atrial contraction (PAC) activity.
5. The system of claim 1 , the AF index representing an expected time until the onset of atrial fibrillation for the patient.
6. A method for predicting the onset of postoperative atrial fibrillation (AF) from an electrocardiogram (ECG) signal representing a patient, comprising:
determining parameters representing the activity of the heart of the patient from the ECG signal, wherein determining parameters representing the activity of the heart comprises detecting P-waves within the ECG signal by determining a wavelet transform of the ECG signal, selecting a scale for the wavelet transform that contains the maximum P-wave energy, removing irregular beats and arrhythmic segments of the ECG signal, and verifying the presence of two maxima points in a selected segment of the wavelet transformed signal in the selected scale prior to an identified QRS wave to confirm the presence of a P-wave;
extracting a plurality of features from the determined parameters, the plurality of features comprising at least one feature extracted from the detected P-waves within the ECG signal; and
determining an AF index for the patient, representing the likelihood that the patient will experience AF, from the calculated plurality of features.
7. The method of claim 6 , wherein determining an AF index for the patient comprises providing the extracted plurality of features as inputs to a neural network classifier.
8. The method of claim 6 , wherein extracting a plurality of features comprises extracting at least one feature representing heart rate variability in the time domain.