Instrumentation and methods for intracardiac catheter tracking: Application of ultrasonic ranging and similarity structure analysis

Introduction. Radiofrequency ablation is an effective and potentially curative therapy for a number of well-defined supraventricular arrhythmias. However, the application of ablation techniques in more complex arrhythmias, such as atrial fibrillation and flutter, have been limited by difficulties in creating continuous lines of block via transcatheter ablation lesions. Fluoroscopy is inadequate for determining the real-time, three-dimensional (3-D) positions of intracardiac catheters, and prolonged exposure to ionizing radiation poses a significant health risk for both the patient and medical staff. There is a need to provide, at a minimum, supplemental 3-D catheter position information to assist the physician in navigating the catheter to specific endocardial sites. In the past five years, several 3-D catheter tracking technologies have emerged with the promise of improving intracardiac catheter navigation, and enabling the creation of complex lesion patterns. This dissertation introduces a catheter navigation and visualization method that is based on transit-time ultrasonic ranging and the statistical method of similarity structure analysis.; Methods. Computational methods are developed to reconstruct the relative 3-D positions of multiple catheter-based ultrasonic transducers from intercatheter range data. Electrode positions along the catheters are computed using a parametric cubic spline. A geometric model of multiple intracardiac electrophysiology catheters is generated using a formalin-fixed ovine heart and a 3-D spatial digitizer. The geometric model is used in mathematical simulations to determine how accurately the catheter tracking method can reconstruct an array of catheter transducers and electrodes under controlled conditions. A state-of-the-art ranging system is designed and implemented using a high speed digital signal processor core to perform gain control and matched filter signal detection functions. In vitro tracking of a roving electrophysiology catheter was performed in a water tank under low noise conditions with a stationary frame of reference. A in vivo pilot study is performed to assess the tracking system performance under more rigorous conditions.; Results. The simulation results show that a catheter tip positional accuracy of 2 mm can be achieved if the intertransducer distances are measured to within ±2%. Comparison of the positional accuracy at each transducer and electrode showed a high sensitivity to the spatial distribution of the transducers. This suggests a potential improvement in accuracy with prudent positioning of the reference catheters or use of additional transducers. Reduction in the distance measurement error provides a propagating improvement in the transducer and electrode positional accuracies. The lower distance measurement errors not only improve the coordinate fit from the multidimensional scaling algorithm, but also reduce misalignment errors from the Procrustes similarity transform. In vitro tracking in a water tank using a static frame-of-reference showed a mean absolute distance error of −0.322 ± 0.468 mm, when moving the roving catheter in 20 mm increments. An in vivo pilot study showed promising results; however, additional optimization of the signal detection...