AT/AF Episode Detection - ICM Feature
AT/AF Episode Detection is a Medtronic Insertable Cardiac Monitor (ICM) feature that detects regular and irregular atrial arrhythmias.
This feature is appropriate for:
- Patients experiencing palpitations and rapid heartbeats associated with atrial arrhythmias.
- Monitoring for the recurrence of atrial arrhythmias in patients who are post atrial ablation.
- Monitoring AT/AF burden and the occurrence of asymptomatic episodes of AT/AF, to assess whether medical treatment is necessary or should be adjusted.
This feature can be found in some Medtronic insertable cardiac monitor (ICM) devices. Please go to manuals.medtronic.com or consult with your local Medtronic representative regarding device models available in your geography.
All patient and clinical data on the screenshots in this document are fictitious and for demonstration purposes only.
Reveal LINQ™ and LINQ II™ ICM
Nominals are set based on the selected Reason for Monitoring.
|Reason for Monitoring||AF Detection Threshold||Ectopy Rejection||Episode Storage Threshold|
LINQ II: ≥ 10 minutes
Reveal LINQ: LONGEST ONLY
|Seizures||Least||Aggressive||≥ 10 minutes|
|Ventricular Tachycardia||Least||Aggressive||≥ 10 minutes|
|Palpitations||Less||Nominal||≥ 6 minutes|
|Suspected AF||Less||Nominal||≥ 6 minutes|
|Other||Less||Aggressive||≥ 10 minutes|
The automatic detection and ECG storage of AT/AF episodes is turned on when you activate Device Data Collection.
Reveal DX/XT, Reveal LINQ, LINQ II
AT/AF Episode Detection identifies the occurrence of AT/AF episodes from variations in the ventricular rhythm. AT/AF episodes are detected using an automatic algorithm based on the pattern of R-wave interval variability within 2-minute periods. The differences between consecutive R-wave intervals are plotted in a Lorenz plot (scatter plot) in order to reveal patterns in the rhythm. Pattern recognition, derived from the Lorenz plot is then used to identify AT and AF episodes. R-wave intervals during AF episodes are highly irregular and uncorrelated, whereas more regular R-wave patterns are expected during AT episodes and sinus rhythm. A clinical AT rhythm with some irregularity may be classified as AF.
The difference between 2 consecutive R-wave intervals (ΔRRn) is plotted on the vertical axis in the Lorenz plot. The preceding R-wave interval difference (ΔRRn-1) is plotted on the horizontal axis. Below are some examples of Lorenz plots showing the different patterns used to identify AT and AF episodes. Note the pattern of scattered dots shown in the AF episode Lorenz plot illustrating irregular R-R intervals. The pattern of more regular R-R intervals displayed on the normal sinus rhythm Lorenz plot shows a pattern with the dots close together instead of scattered.
Lorenz Plot Showing AF
Lorenz Plot Showing AT
Lorenz Plot Showing Normal Sinus Rhythm
AF detection settings are auto-adjusted based on the Reason for Monitoring, which is found in Device Data Collection. The nominal settings for AF Only detection ensure optimal performance of AF detection in most patients. In some instances, the parameter may be reprogrammed to suit individual patients.
AF Only Detection: Sensitivity
The Sensitivity setting is not related to sensing of R-waves, but rather the sensitivity of R-R irregularity, or the threshold to declare AF. If in the 2 minutes of analysis, the Lorenz plot calculations conclude greater irregularity than the programmed sensitivity, AF is declared for those 2 minutes.
The options for programming the AT/AF Detection sensitivity are:
- Least Sensitive
- Less Sensitive
- Balanced Sensitivity
- More Sensitive
For example, if sensitivity is programmed to Least Sensitive, then the AF threshold is at the highest level, which means that AF is detected when the ventricular rhythm has more irregularity.
An example of the AF calculation based upon the sensitivity setting is shown in the figure below.
A primary cause of false positive AF detection is a run of ectopy (PACs or PVCs) with irregular coupling of intervals caused by underlying sinus variability. Two rejection algorithms exist to prevent false detections.
- The Ectopy Rejection algorithm recognizes patterns of ectopy by the density of points in the Lorenz Plot.
- The P-wave presence algorithm looks for evidence of a P-wave between two R-waves.
The options for programming Ectopy Rejection are:
- Nominal (P-wave presence )
- Aggressive (Ectopy Rejection + P-wave presence)
When Ectopy Rejection is set to "Nominal" the P-wave presence algorithm is enabled. When Ectopy Rejection is set to "Aggressive", both the P-wave presence and the Ectopy Rejection algorithms are enabled. The device will not detect AF when the algorithm detects evidence of ectopy during any 2 min period. When Ectopy Rejection is set to "Off", both the P-wave presence and the Ectopy Rejection algorithms are disabled.
Examples of AF calculations are given in the figures below for Ectopy Rejection set to "Nominal" and "Aggressive" modes.
There are limitations associated with the P-wave evidence algorithm including:
the P-waves become too small to sense
the P-waves are masked by a noisy baseline
the patient’s heart rate is greater than 77 bpm and the T-waves may overlap with the P-wave window.
PSENSE (P-wave Presence Algorithm)
Reveal LINQ, LINQ II
The Reveal LINQ and LINQ II ICMs have an algorithm that helps discriminate between true and false AF target to improve the specificity for patients with Sick Sinus Rhythm or PACs. As mentioned earlier, AF evidence analyzes R-R variability over 2-minute periods. P-sense uses the same 2-minute interval to monitor for P-wave morphology and builds evidence of the presence of P-waves. The AF evidence must be above the programmed threshold to be considered AF. When there is a high degree of R-R variability, yet sufficient evidence of P-waves the Modified AF Evidence rejects false episodes when P-waves are present.
AF EVIDENCE – P-WAVE EVIDENCE = MODIFIED AF EVIDENCE
Reveal LINQ & LINQ II
In TruRhythm™ Detection, a self-learning algorithm has been added to the Reveal LINQ and LINQ II ICM to enhance the AT/AF Episode Detection feature. The TruRhythm self-learning algorithm keeps a history of P-wave evidence and uses this history to adapt the level of AF evidence needed to detect an AF episode, thereby reducing the number of false positive AF episode detected in patients with Sick Sinus Syndrome or PACs. As mentioned earlier, AF evidence analyzes R-R variability over 2-minute periods. P-sense uses the same 2-minute interval to monitor for adequately sensed P-waves and builds evidence of the presence of P-waves. The AF evidence must be above the programmed threshold to be considered AF. When there is a high degree of R-R variability, yet sufficient evidence of P-waves, TruRhythm detection will correctly reject these episodes as AF.
AF EVIDENCE – P-WAVE EVIDENCE = MODIFIED AF EVIDENCE
When high evidence of P-waves is found, the Modified AF Evidence is used for the rest of the day. It does not start over for the next 2-minute period. The self-learning algorithm allows the device to remember and adapt the P-wave evidence throughout the day. LINQ™ ICMs with TruRhythm Detection may be identified by the serial number. In addition, you will see “LINQ™ + TruRhythm™” on the pdf heading of all CareLink™ Reports. Consult with your local Medtronic Representative on the changes specific to your geography.
Self-Learning Algorithm Summary
Reveal LINQ, LINQ II
The AF algorithm tracks R-wave variability and collects the P-wave evidence history.
This P-wave evidence is used to adapt the AF detection threshold.
This adaptation was designed to improve the rejection of false AF episodes.
If you select AT/AF as the detection Type, the following programming screen appears. Nominal settings are displayed.
Note: Program AT/AF Detection to "Off" to prevent automatic detection of AT/AF episodes.
Detect Very Regular AT Rhythms
Detection of very regular AT rhythms may be programmed. This feature discriminates AT episodes by analyzing the regularity of the patient's heart rate. The four programming options are:
- On – Rate >= 67 bpm
- On – All Rates
If the heart rate during an AT episode is less than the programmed "Detect Very Regular AT Rhythms" parameter setting, the AT Episode is not detected.
If the "Detect Very Regular AT Rhythms" parameter setting is set to "On – All Rates", the system detects all AT episodes.
If the "Detect Very Regular AT Rhythms" parameter setting is set to "Off", very regularly conducted atrial tachycardias such as consistent 2:1 or 3:1 atrial flutter will not be detected.
Note: An AT/AF episode can occur simultaneously with one of the other types of episodes. If this occurs and detection is programmed On, the Reveal LINQ device will store information and an ECG recording for both types of episodes.
AT/AF Recording Threshold
Arrhythmia episode data is stored in an episode log. This includes up to 30 episodes each for AT, and AF. For each AT/AF episode the device stores an ECG of the first 2 min of the episode. The AT/AF Recording Threshold allows for recording to be enabled based upon episode length.
There are several programmable options for AT/AF Recording Threshold:
- All episodes
- Episodes ≥ 6, 10, 20, 30 or 60 mins
- Only Longest Episode (≥ 10 min)
If AT/AF Recording Threshold is programmed to Only Longest Episode, an AT/AF episode of 10 min or greater duration is required to store an episode log and ECG. Until the device is interrogated, each subsequent AT/AF episode that is of greater duration overwrites the previously recorded shorter episode.
Program AT/AF detection to “AF Only” for AF detection, “AT/AF” to include AT detection or “OFF” to prevent automatic detection of AT/AF episodes.
Select an AF duration that would closely match when clinical action would be taken
Example: if following a patient for suspicion of AF post stroke 2min of AF may be sufficient to consider drug treatment. However, if no action would be taken until there were 1 hour of AF, then a 30 minute or 60-minute duration may be more appropriate.
When ectopy is present consider setting Ectopy Rejection to “Aggressive”
Select the AF Detection Sensitivity based on desired sensitivity. If false episodes are observed, consider setting AF Detection Sensitivity to a less sensitive setting (ie: BALANCED -> LESS).
An AT or AF episode CareAlert™ notification may be enabled monitoring Reveal LINQ™ and LINQ II™ ICM patients remotely.
- Reveal LINQ
The Reveal LINQ ICM sends a daily wireless audit transmission to the patient's MyCareLink™ monitor. When an AF episode is detected and transmitted to CareLink and the CareAlert notification for an AF episode is enabled on the CareLink® website, a CareAlert Notification will be generated. Event details will be sent to CareLink network in the daily wireless audit and displayed on the Event Report. The episode details include 30 seconds of the detected event's ECG if its the highest priority event of the day else 10sec of ECG from the longest AF episode of the day. In addition to the ECG data, 2 minutes of heart rate plot will be included. The clinician receives the CareAlert notification based on clinic defined methods.
- LINQ II
The LINQ II ICM sends data to the patient's monitor (MyCareLink Relay™ or MyCareLink Heart™ mobile app) throughout the day. When an AT or AF episode is detected and transmitted to CareLink and the CareAlert notification for a AT or AF episode is enabled on the CareLink® website, a CareAlert Notification will be generated. CareAlerts are processed between 5-6am based on the local clinic time zone. The clinician receives the CareAlert notification based on clinic defined methods.
Stored episodes can be viewed remotely on the CareLink™ network by navigating to the Episodes tab under Transmission Details. For in-person follow-up stored episodes can be viewed using the 2090 Programmer (Reveal DX/XT and Reveal LINQ only) and LINQ Mobile Manager (LMM) by opening Episodes and selecting from the Arrhythmia Episodes log.
Reveal DX/XT, Reveal LINQ
Up to 13 AT/AF episodes with ECG will be stored when AT/AF detection is programmed "On". For each AT/AF episode, the ECG recording consists of 2 minutes of ECG prior to initial detection. When memory is full, data from the most recent AT/AF episode may overwrite the oldest stored AT/AF or the oldest auto detected episode data. In addition to the episodes available with ECG up to 30 logs for each type could be stored.
Up to 6 AT/AF episodes, 3 AF and 3 AT, with ECG and up to 7 additional episodes for each type without ECG (log only) will be stored per day when AT/AF detection is programmed "On". The 3 episodes for each type will be as follows: The first 2 and the longest episode of the day (24hour day beginning at 12am). For each AT/AF episode, the ECG recording consists of 2 minutes of ECG prior to initial detection. When memory is full, data from the most recent AT/AF episode may overwrite the oldest stored AT/AF or the oldest auto detected episode data. In addition to the daily episode storage up to 30 logs for each type could be stored.
The log includes the following summary information:
- Type of episode – this will say "AF" for an AF event.
- The date, time, and duration of the episode.
- The highest ventricular rate.
- The median ventricular rate at detection.
- Whether ECG data is available for the episode.
For a particular episode, you can display the following information:
- An interval (or rate) plot
- A strip chart of the stored ECG (if available)
- A text summary
Marker channels for AT/AF episodes are:
FD (above baseline): Initial detection of an AF episode
TD (above baseline): Initial detection of an AT episode
Note: If the patient uses the Patient Assistant while an automatically detected episode is in progress, both episodes are stored. The device records "Symptom (Patient Activated) occurred during episode" in the text of the automatically detected episode.
The examples below illustrate an event report that was generated by an AF CareAlert™ notification. In both examples, the interval plot clearly depicts the variability in the R-R intervals that led to AF episode detection. The episode details are also displayed. The first example seen below is page 1 of a Reveal LINQ ICM with TruRhythm Episode report. Notice that “Reveal LINQ™ + TruRhythm™” is displayed at the top of the report and the length of the ECG strip is 30 seconds. A VS marker corresponds with each QRS complex and AF Detection is observed on the ECG.
Example 1: Reveal LINQ + TruRhythm Event Report
The second example seen below is page 1 of a Reveal LINQ Event Report and the ECG strip is 10 seconds in length. A VS marker corresponds with each QRS complex and AF Detection is observed on the ECG.
Example 2: Reveal LINQ Event Report
Sources: Medtronic Reveal LINQ™ LNQ11 Clinician Manual. Medtronic LINQ II™ LNQ22 Clinician Manual. TruRhythm™ Detection Algorithms. Medtronic data on file. 2017.