Description of the proposed use of animals
Animals will be used to study the effectiveness of a new imaging mode to characterize ablation lesions. Acoustic Radiation Force Impulse (ARFI) imaging has been shown by us to differentiate normal from ablated myocardium. This imaging modality has been implemented by us and shown at our institution in animals to be capable of in vivo characterization of ablation lesions. In order to utilize this technology, we must integrate it into an electranotomical mapping system. This integrated system must also be tested for image registration accuracy in order to guide ablation therapy.
Justification for the use of animals
We commonly use tissue phantoms and tissue samples obtained from qualified sources to create lesions and image their properties. We also commonly use computer models to predict lesion size and test various aspects of ARFI imaging signal processing. However, the creation and imaging of a lesion in vivo can not be simulated with current technology. Thus, animal testing is required prior clinical trials of this promising new technology.
We have requested 12 animals in each of the last two years of the proposal. These animals will be used in two proposed studies, one to characterize the accuracy of the proposed multi-modal imaging system. We are assuming that the position error in both systems will be small with a mean value of less than 1 mm. Assuming a mean distance error of 1 mm and a SD of 0.25 mm (within animal var .0625 mm2). With these assumed errors, there is a 8% chance of making a TYPE II error when α=.05 on a one tailed (distance error is positive only) test and a sample size of 12.
It is more difficult to calculate a power for the contingency table analysis. However, if we assume 100 lesions per animal, and that the frequency counts will all be fairly close to 1200, that is all of the methods will be good at detecting block in these experiments. In this case, the sensitivity of the analysis to a single change in outcome will be less than 0.1% with 12 animals. For a 5% error, 50 lesions will have to be graded incorrectly or assigned to the wrong category. We consider this an acceptable range for error.