Publication detail

Analysis of Dynamic Changes in ECG Signals During Optical Mapping By Dynamic Time Warping

VYKLICKÝ, M. KOLÁŘOVÁ, J. PROVAZNÍK, I. NOVÁKOVÁ, M. BLAHA, M.

Original Title

Analysis of Dynamic Changes in ECG Signals During Optical Mapping By Dynamic Time Warping

English Title

Analysis of Dynamic Changes in ECG Signals During Optical Mapping By Dynamic Time Warping

Type

conference paper

Language

en

Original Abstract

Voltage-sensitive dyes (VSDs) are used in touch-less recording of electrical activity from animal hearts. VSD's undergo changes in their fluorescence spectra, in response to changes in the surrounding electric field. This optical response is sufficiently fast to detect transient potential changes in excitable cardiac cells. However, our previous Langendorf perfused heart studies have suggested changes in ECG signals in time and time-frequency domain caused by application of VSD during optical mapping. Detailed analysis of changes is needed to reveal impact to particular fractions of a heart cycle. Rabbit hearts are employed in our laboratory to record monophasic action potentials using VSDs in isolated hearts perfused according to Langendorf. The spontaneously beating heart is mounted on Langendorf set and perfused with constant perfusion pressure with Krebs-Henseleit solution (37 C, 1,25mM Ca2+). After control period (20 min), the heart is loaded with di-4-ANEPPS (2uM) diluted in perfusion solution (25 min) and then washed-out with dye-free perfusion solution (25 min). ECG signals are recorded from three bipolar leads positioned around the heart. Signals are amplified and digitized at sampling rate of 4000 Hz by a 16-bit AD converter. Dynamic time warping (DTW) is applied to PQ segments, QRS complexes, and ST segments of selected consecutive heart cycles in each experiment period. An optimal path resulting from DTW based on minimum distance between the control segments (before the application VSD) and segments recorded in next two experiment steps is computed. It has been observed that during the loading period the changes in the optimal path exceeded 50% of its length in all three analyzed segments in all experiments. In the end of wash-out period, considerable changes in the optimal path were presented in QRS complexes (52%), lower level of changes were in ST segments (31%), and minor changes were found in PQ segments (17%). Thus, it may be concluded that the optimal path of dynamic time warping applied to ECG signals is sensitive to various parts of optical mapping procedure. Moreover, DTW was applied to analysis of selected fractions of a heart cycle and is proper for analysis of VSD influence to conductive system of the heart.

English abstract

Voltage-sensitive dyes (VSDs) are used in touch-less recording of electrical activity from animal hearts. VSD's undergo changes in their fluorescence spectra, in response to changes in the surrounding electric field. This optical response is sufficiently fast to detect transient potential changes in excitable cardiac cells. However, our previous Langendorf perfused heart studies have suggested changes in ECG signals in time and time-frequency domain caused by application of VSD during optical mapping. Detailed analysis of changes is needed to reveal impact to particular fractions of a heart cycle. Rabbit hearts are employed in our laboratory to record monophasic action potentials using VSDs in isolated hearts perfused according to Langendorf. The spontaneously beating heart is mounted on Langendorf set and perfused with constant perfusion pressure with Krebs-Henseleit solution (37 C, 1,25mM Ca2+). After control period (20 min), the heart is loaded with di-4-ANEPPS (2uM) diluted in perfusion solution (25 min) and then washed-out with dye-free perfusion solution (25 min). ECG signals are recorded from three bipolar leads positioned around the heart. Signals are amplified and digitized at sampling rate of 4000 Hz by a 16-bit AD converter. Dynamic time warping (DTW) is applied to PQ segments, QRS complexes, and ST segments of selected consecutive heart cycles in each experiment period. An optimal path resulting from DTW based on minimum distance between the control segments (before the application VSD) and segments recorded in next two experiment steps is computed. It has been observed that during the loading period the changes in the optimal path exceeded 50% of its length in all three analyzed segments in all experiments. In the end of wash-out period, considerable changes in the optimal path were presented in QRS complexes (52%), lower level of changes were in ST segments (31%), and minor changes were found in PQ segments (17%). Thus, it may be concluded that the optimal path of dynamic time warping applied to ECG signals is sensitive to various parts of optical mapping procedure. Moreover, DTW was applied to analysis of selected fractions of a heart cycle and is proper for analysis of VSD influence to conductive system of the heart.

Keywords

ECG signal, optical mapping, dynamic time warping

RIV year

2005

Released

02.09.2005

Publisher

IEEE

Location

Francie

ISBN

0-7803-9337-6

Book

Computers in Cardiology

Pages from

543

Pages to

546

Pages count

4

BibTex


@inproceedings{BUT17862,
  author="Marek {Vyklický} and Jana {Kolářová} and Ivo {Provazník} and Marie {Nováková} and Milan {Blaha}",
  title="Analysis of Dynamic Changes in ECG Signals During Optical Mapping By Dynamic Time Warping",
  annote="Voltage-sensitive dyes (VSDs) are used in touch-less recording of electrical activity from animal hearts. VSD's undergo changes in their fluorescence spectra, in response to changes in the surrounding electric field. This optical response is sufficiently fast to detect transient potential changes in excitable cardiac cells. However, our previous Langendorf perfused heart studies have suggested changes in ECG signals in time and time-frequency domain caused by application of VSD during optical mapping. Detailed analysis of changes is needed to reveal impact to particular fractions of a heart cycle.

Rabbit hearts are employed in our laboratory to record monophasic action potentials using VSDs in isolated hearts perfused according to Langendorf. The spontaneously beating heart is mounted on Langendorf set and perfused with constant perfusion pressure with Krebs-Henseleit solution (37 C, 1,25mM Ca2+). After control period (20 min), the heart is loaded with di-4-ANEPPS (2uM) diluted in perfusion solution (25 min) and then washed-out with dye-free perfusion solution (25 min). ECG signals are recorded from three bipolar leads positioned around the heart. Signals are amplified and digitized at sampling rate of 4000 Hz by a 16-bit AD converter. Dynamic time warping (DTW) is applied to PQ segments, QRS complexes, and ST segments of selected consecutive heart cycles in each experiment period. An optimal path resulting from DTW based on minimum distance between the control segments (before the application VSD) and segments recorded in next two experiment steps is computed.

It has been observed that during the loading period the changes in the optimal path exceeded 50% of its length in all three analyzed segments in all experiments. In the end of wash-out period, considerable changes in the optimal path were presented in QRS complexes (52%), lower level of changes were in ST segments (31%), and minor changes were found in PQ segments (17%).

Thus, it may be concluded that the optimal path of dynamic time warping applied to ECG signals is sensitive to various parts of optical mapping procedure. Moreover, DTW was applied to analysis of selected fractions of a heart cycle and is proper for analysis of VSD influence to conductive system of the heart.",
  address="IEEE",
  booktitle="Computers in Cardiology",
  chapter="17862",
  institution="IEEE",
  year="2005",
  month="september",
  pages="543--546",
  publisher="IEEE",
  type="conference paper"
}