The objective is to relate extensively sampled distributions of electrocardiographic signals (maps) to local cardiac excitation and recovery.
Specific aims are to 1) determine accuracy of estimating epicardial excitation and recovery from body surface maps, 2) evaluate estimates of endocardial events from intracavitary maps, 3) define features indicative of myocardial lesions in the presence of left bundle branch block (LBBB), 4) determine the physiologic basis of statistically derived map features, and 5) improve the solution to forward and inverse ECG problems by correlation of information from endocardial and epicardial surfaces. Special features of methodology are our instrumentation for acquiring and computer techniques for processing multilead ECG data.
For aims 1 and 2, QRS and T derivatives and distributions of deflection areas will be evaluated as markers of excitation and recovery.
For aim 3, LBBB will be simulated by right ventricular pacing and myocardial lesions produced.
For aim 4, a variety of electrophysiologic states will be produced to provide maps from which orthogonal, spatial features are determined by the Karhunen-Loeve analysis. A hierarchy of complexity of features will be correlated with simplicity of cardiac sources.
For aim 5, the ability of multisurface ECG data to improve accuracy of forward and inverse solutions will be tested by experiment and simulation. Significance of the project is that heart disease is often localized and the extensive sampling of maps provides a locally selective examination likely to improve diagnosis and reduce need for high cost/risk tests. Examples of utility include recognition of arrhythmia vulnerable states without invasive electrophysiologic testing and recognition of myocardial disease in the presence of LBBB.
Showing the most recent 10 out of 17 publications
