Automatic Detection of Larval Zebrafish ECG: Computational Tool for High-throughput Cardiac Activity Analysis

Abstract

Automatic analysis of larval zebrafish electrocardiographs (ECG) is essential for high-throughput measurements in environmental toxicity assays, cardiotoxicity measurements and drug screening. We have developed a MATLAB based software is built on methods that have previously been used to analyze human ECG, such as the Pan-Tompkins algorithm and Wavelet. For the first time these sophisticated tools have been applied to larval zebrafish ECG to automatically characterize the heart-beat waveforms. The ability of the automated algorithm to detect the QT interval for both normal and pharmacologically altered larval ECG is found and compared to previously used software that is built into LabChart(R) (AD Instruments). Using wavelet transforms it is shown that the typical larval ECG features are within the frequency range of 1 to 31 Hz. It is also shown that the automated software is capable of detecting QTc (heartrate corrected heartbeat interval) changes within pharmacologically altered zebrafish larval ECG. The automated process is a significant improvement on the approaches that were previously applied to the zebrafish ECG. The automated process described here that is based on established techniques of analyzing ECG can sensitively measure pharmacologically induced changes in the ECG. The novel, automated software is faster, more sensitive at detecting ECG changes and less dependent on user involvement, thus minimizing user error and human bias. The automated process can also be applied to human ECG.