Catheter ablation is a medical procedure that involves the destruction of small volumes of heart tissue with radiofrequency energy. To be successful, catheter ablation requires precise localization of the tissue to be destroyed. To accomplish this in a typical ablation procedure, five catheters containing a total of 19 electrode pairs are utilized to record potentials (called electrograms) from spatially distinct locations within the heart throughout the cardiac cycle. Experienced cardiologists interpret the electrograms to locate the conduction pathways responsible for the arrhythmia. The pathways are destroyed by applications of radiofrequency energy, thus treating the arrhythmia. Ablation procedures are performed by highly trained and experienced cardiology sub-specialists yet the massive amount of data produced during these procedures creates a data overload problem that can impede the performance of even the best practitioners. This may be evidenced by (1) overlooking important signal features, (2) misinterpreting the signals, and (3) misinterpreting catheter locations in the heart, all of which may lead to increased procedure duration and/or applications of radiofrequency energy to the wrong part of the heart. The purpose of this project is to develop a model-based system for interpreting intracardiac electrograms in near real-time. The system is intended to assist physicians in """"""""making sense"""""""" of the enormous amounts of data recorded during a cardiac electrophysiology study. New computer algorithms for reasoning about the time- and space-varying nature of intracardiac electrograms from underlying causal models of the heart will be developed. The models can be represented as a graph of nodes connected by arcs. The nodes represent specific an atomic regions of the heart while the arcs represent the connections between the regions. The analytic approach will be a variation of the hypothesize-and-test paradigm. The control loop will be based on the """"""""tracking"""""""" concept whereby models will be tracked as long as the data supports the model. Rule-based knowledge will be utilized to generate models based on the observed data. The output of the system will be a series of ladder diagrams describing the data. In the event that the data admits more than a single explanatory model, ladder diagrams will be generated for all created models. This proposal is an extension of the applicant's graduate and post- doctoral project, which used the same approach to the simpler domain of the interpretation of the body-surface electrocardiogram. The domain of this proposal is more complex because it adds reasoning with a more detailed three dimensional cardiac model to the temporal reasoning required for previous work. Also, this domain requires the ability to account for simultaneous activation of the heart at multiple locations, which was performed at a rudimentary level in the previous work. This project is important for three reasons: (1) it offers new knowledge- based algorithms for reasoning about time- and space-varying data, (2) a comprehensive and extensible model of the cardiac conduction will be created that incorporates the spatial resolution necessary for interpreting intracardiac electrograms, and (3) a software tool such the one proposed may help clinicians improve the quality of health care.

Agency
National Institute of Health (NIH)
Institute
National Library of Medicine (NLM)
Type
First Independent Research Support & Transition (FIRST) Awards (R29)
Project #
7R29LM006004-02
Application #
2032386
Study Section
Biomedical Library and Informatics Review Committee (BLR)
Project Start
1995-12-15
Project End
2000-11-30
Budget Start
1996-12-01
Budget End
1997-11-30
Support Year
2
Fiscal Year
1997
Total Cost
Indirect Cost
Name
University of Texas Health Science Center San Antonio
Department
Internal Medicine/Medicine
Type
Schools of Medicine
DUNS #
800772162
City
San Antonio
State
TX
Country
United States
Zip Code
78229