부산대학교 도서관

This paper presents an approach to surpass the common linearization of electrical cardiac signals and to use signal components of high frequencies for clinical examinations. For this, a high-resolution data acquisition equipment was designed. On-line real time segmentation as well as off-line evaluation of endocardial signals were carried out by wavelet packet transformation. In order to ensure the connection to the signal-generating process a modified Luo-Rudy model of the action potential and a model for two-dimensional spreading of excitation were developed. We found characteristic subbands in the range from 3 kHz to 5 kHz in each of our first 25 recordings. The method is suitable to determine direction and speed of the endocardial excitation front. Thus, it is useful for endocardial mapping and ablation technique. We are currently carrying out extended clinical studies.