Analyzing Long-Term Electrocardiography Recordings to Detect Arrhythmias in Mice

Analysis of large ECG data sets, especially when you screen for any arrhythmia, is very challenging and labor intensive. Our semi-automated approach allows for much faster and more convenient screening for arrhythmias in mice. This semi-automated approach leads to more accurate results.

The procedure will be demonstrated by Aparna Chivukula, a PhD student in my lab. Begin by analyzing basic ECG parameters. Click on Subject Setup or Channel Details to select any of the derived parameters.

As mice follow a circadian rhythm, analyze two resting ECGs per day. One at daytime and one at nighttime to control for circadian effects, selecting suitable times depending on the light on or off cycle in the animal facility. Select a section of the ECG with good signal quality and a stable heart rate in the heart rate trend graph within a defined reasonable timeframe around this time point.

Confirm the accuracy of the validation masks, or adjust manually to 20 consecutive QRS complexes, and add missing validation marks. For further calculations and visualizations, mark the lines containing the values of these 20 consecutive QRS complexes in the derived parameter list, and copy to a spreadsheet or statistics software. Open data insights by clicking on Experiment Data Insights and observe the search panel at the top.

Then display the Results panel. Observe the number of search hits displayed as a histogram in the bottom of the panel. Given that the normal heart rate of a mouse is 500 to 724 beats per minute, define a search rule, Bradycardia, to detect bradycardia.

Right-click within the search list and select Create New Search to open the search entry dialogue. Right-click within the white box and select Add New Clause. Using the dropdown menus and text fields, define the search rule, Bradycardia Single, as value less than 500, and click on Okay to add this search rule to the list.

Apply this search rule by clicking and dragging it to the channel of interest on the left. As bradycardia requires more than one long RR interval, define an additional search rule Bradycardia, to account for this. Click on Okay to add this search rule to the list.

Apply this search rule by clicking and dragging it to the channel of interest on the left. To confirm bradycardia, and to reject false results, review each result manually by left-clicking on the waveform. Press STRG plus R to reject the selected result, which will then disappear from the list.

To define a search rule to detect tachycardia, right-click within the search list and select Create New Search to open the search entry dialogue. Then right-click within the white box and select Add New Clause. Using the dropdown menus and text fields, define the search rule, Tachycardia Single, as value greater than 724, and click on Okay to add this search rule to the list.

Apply this search rule by clicking and dragging it to the channel of interest to the left. As tachycardia requires more than one short RR interval, define an additional search rule, Tachycardia, to account for this, and click on Okay to add this search rule to the list. Apply this search rule by clicking and dragging it to the channel of interest on the left.

To confirm tachycardia, and to reject false results, review each result manually by left-clicking on the waveform and use the shortcut STRG plus R to reject the selected result. To detect sinoatrial and atrioventricular blocks, right-click within the search list and select Create New Search to open the search entry dialogue. Then right-click within the white panel and select Add New Clause.

Using the dropdown menus and text fields, define the search rule, Pause, as value greater than 300, and click on Okay to add this search rule to the list apply this search rule by clicking and dragging it to the channel of interest to the left. To confirm a pause, decide if the pause is a sinoatrial or atrio ventricular block, and reject false results. Review each result manually by left-clicking on the waveform and press STRG plus R to reject the selected result, which will disappear from the list of results.

This approach allowed the detection of periods of reduced heart rate, increased atrioventricular and ventricular conduction, as well as altered repolarization in mice treated with a new drug. Furthermore, the protocol was used to determine the time course of arrhythmia occurrence in the form of traces represented at normal sinus rhythm, sinus pause, three degrees of atrioventricular block, and atrial fibrillation. Final readouts are manifold and depend on study design and research topic.

In atrial fibrillation, total disease burden has been established as a prognostic marker and could be calculated from data obtained with this protocol.