Frequency analysis for predicting left ventricular dysfunction
AssigneeThe Cleveland Clinic Foundation
InventorMark J. Niebauer
Life of this patent
Estimating the term…
Abstract
Systems and methods are provided for evaluating infranodal pacing is applied to a patient. Electrocardiogram (ECG) data representing the pacing is obtained from a set of electrodes as an ECG lead. A predictor value representing a frequency content a portion of the ECG lead is extracted. A fitness parameter is determined for the pacing from at least the predictor value. The fitness parameter represents a likelihood that the applied infranodal pacing will induce left ventricular dysfunction in the patient. The fitness parameter is displayed to a user at an associated display.
Claims18
A method comprising:
- applying infranodal pacing to a patient at a pacing site;
- obtaining electrocardiogram (ECG) data representing the pacing site from a set of electrodes as an ECG lead;
- extracting a predictor value representing a frequency content of a selected time window of the ECG lead, the predictor value for the frequency content of the selected time window of the ECG lead comprising a change in the median frequency across the selected time window of the ECG lead, wherein the change in the median frequency is determined by dividing the selected time window of the ECG lead into a plurality of subregions, determining a median frequency for each subregion, and calculating one of a slope of the median frequency across the plurality of subregions and a difference between the median frequency of a first selected subregion and a second selected subregion;
- determining a fitness parameter for the pacing site from at least the predictor value, the fitness parameter representing a likelihood that the infranodal pacing applied at the pacing site will induce left ventricular dysfunction in the patient;
- and
- displaying the fitness parameter to a user at an associated display.
A system comprising:
- a signal processor configured to receive an ECG lead and isolate a selected time window from the ECG lead during infranodal pacing of a patient;
- a predictor extraction component configured to extract a predictor value representing a frequency content of the selected time window, the predictor value for the frequency content of the portion of the ECG lead comprising a change in the median frequency across the selected time window of the ECG lead, wherein the change in the median frequency is determined by dividing the selected time window of the ECG lead into a plurality of subregions, determining a median frequency for each subregion, and calculating one of a slope of the median frequency across the plurality of subregions and a difference between the median frequency of a first selected subregion and a second selected subregion;
- and
- a parameter calculation component configured to determine a parameter representing the likelihood that the patient will experience left ventricular dysfunction induced by the infranodal pacing from at least the predictor value.
A method comprising:
- applying infranodal pacing to a patient using a first pacing method;
- obtaining electrocardiogram (ECG) data representing the first pacing method from a set of electrodes as a first ECG lead;
- extracting a predictor value representing a change in a median frequency across a QRS complex of the first ECG lead, wherein the change in the median frequency is determined by dividing the QRS complex of the ECG lead into a plurality of subregions, determining a median frequency for each subregion, and calculating one of a slope of the median frequency across the plurality of subregions and a difference between the median frequency of a first selected subregion and a second selected subregion;
- determining a fitness parameter for the first pacing method from at least the first predictor value, the fitness parameter representing a likelihood that the infranodal pacing using the first pacing method will induce left ventricular dysfunction in the patient;
- and
- selecting a second pacing method for the patient if the fitness parameter does not meet a threshold value.
Dependencies are inferred from the claim text.