Publication details
Statistical Analysis in Complex-Valued Wavelet Analysis of Voltage-Sensitive Dye Mapping
| Authors | |
|---|---|
| Year of publication | 2007 |
| Type | Article in Proceedings |
| Conference | Computers in Cardiology |
| MU Faculty or unit | |
| Citation | |
| Keywords | complex-valued wavelet analysis;statistical analysis;voltage-sensitive dye |
| Description | At present, the optical mapping is widely used in cardiac electrophysiology animal experiments. The principle of optical mapping is an application of voltage-sensitive dye (VSD) to examined tissue where it binds to a membrane of cardiac cells. It is necessary to examine recordings from all phases of the experiment to exclude possible negative influence of the dye to the tissue. Also, the same procedure could be used to detect electrophysiological changes e.g. during ischemia studies. The continuous wavelet transform (CWT) has been used to detect expected changes. The use of CWT may help to track the changes simultaneously in time and frequency domain. Statistical analysis of CWT results supports hypothesed changes. 15 rabbit hearts were included in the study. Each heart was mounted on a Langendorff apparatus, filled with Krebs-Henseleit (K-H) solution (1.25 mM Ca2+) and placed in a bath (37C). The hearts were perfused at the constant pressure of 85 mmHg. The hearts were stabilized for 15 minutes. After control period, the hearts were perfused with 1mM solution of VSD di-4-ANNEPS diluted in K-H solution (loading period). Long term trends were studied via analysis of 15 minutes of loading period, and 15 minutes of wash-out. ECG signals were continuously recorded during whole experiment. 10-second segments at the beginning of each minute of records were extracted and used. In all phases, heart beats were detected and segmented into P-Q, QRS, and S-T intervals. Segments were transformed by CWT using Gaussian wavelet No.2 to reveal changes in time-frequency domain. Significant abnormalities were assessed by comparing the mean value of each of CWT of the two studied populations by means of a two-way two-tailed t-test to test for the null hypothesis that means are equal. First population was build from ECG segments in control period, second populations from ECG segments in all other periods. Time-frequency domain statistical analysis using modulus and phase of CWT revealed large changes in P-Q segment during VSD loading, lower changes in QRS complex and almost no shape changes in S-T segment. P-Q segment and QRS changes were partly restored in wash-out. These findings corresponds to previous studies based on time-domain analysis but it proved QRS changes and their restoration in wash-out with higher statistical importance. |