Figure 1.
Flowchart representing the workflow of CardioMDA.
On execution of the program, the graphical user interface (GUI) opens up. In phase I, the user loads the data that is to be analysed. The program automatically pre-processes to remove noise and after peak detection, the user is prompted to select a template for analysis. During correlation analysis, the program checks for field potential complexes whose correlation coefficient satisfies the user-defined correlation factor and produces an average waveform. In phase II, the program allows the user to define the onset and offset, either automatically using the built-in algorithms, or manually by the user. The process is repeated for upto ten signals and the field potential duration and other parameters are compared. The results can be saved for future reference.
Figure 2.
Screenshot of the graphical user interface for CardioMDA.
The screenshot presents the Load panel of the CardioMDA software with its interactive interface options and basic signal information. Once the template has been selected, the user then moves to the Average panel to set and execute the correlation analysis. Compare, Analyze and Statistics panel are used to measure the field potential duration and compare it with other drug concentrations.
Figure 3.
Unstable signal morphology as a result of drug treatment.
Signal showing changing morphology as a result of drug treatment. Field potential complex I and field potential complex II have different signal properties in terms of amplitude and duration, thereby accentuating the need for correlation analysis.
Figure 4.
Comparison of correlation analysis results with different values of correlation factor.
An illustration of the correlation analysis procedure for morphology changing (top row) and arrhythmogenic signals showing premature activation (bottom row) for different correlation factors is shown. The field potential complexes in green have a correlation coefficient greater than correlation factor and are identified as true field potential complexes (accepted for averaging) whereas those in red are identified as true negatives (rejected for averaging). The black line presents the morphology of averaged signal. In both cases, it can be seen that for lower value of correlation factor (0.93), a number of false positives occur, causing a significant change in the morphology of the averaged waveform. From the bottom row, it can be seen that a higher correlation factor (0.98) is more accurate in identifying the true field potential complexes.
Table 1.
Cross correlation results for normal, arrhythmogenic and morphology changing signals with different values of correlation factors.
Table 2.
Statistical measure of performance in accurately detecting true field potential complexes in arrhythmogenic and morphology changing signals using cross correlation with CF = 0.98.
Figure 5.
Illustration of smoothing effect after ensemble averaging.
N denotes the number of ensembles used for averaging. The resulting waveform from ensemble averaging is smoother when the number of ensembles being averaged is higher.
Table 3.
Field potential duration prolongation with increased hERG channel blocker E-4031 drug concentrations.
Figure 6.
Overlay plot showing the effect of hERG potassium channel blocker E-4031 on field potential duration.
As a result of E-4031 treatment, the increase in field potential duration is dependent on the concentration of the drug. The effect of the drug can most prominently be seen at the repolarisation wave peak as a change of signal amplitude and the prolongation of the repolarisation wave peak compared to baseline signal. The effect of the drug can also be seen as a prolongation of the offset point at the end of the field potential cycle.
Figure 7.
Overlay plot showing the effect of sotalol on field potential duration and signal shape.
As a result of drug treatment, the repolarisation wave peak, which corresponds ECG T-wave, is prolonged due to increasing concentration of Sotalol. However, the offset point of field potential duration is similar between different Sotalol concentrations.
Figure 8.
Overlay plot showing effect of IKs-blocker JNJ303 on field potential duration and signal shape.
Similar to Sotalol, JNJ303 prolongs the repolarisation wave peak but the field potential duration offset point is similar between different JNJ303 concentrations.