false
Catalog
Mapping Atrial Fibrillation in 2020: Key Updates, ...
Contact Activation Mapping: Global or Simultaneous ...
Contact Activation Mapping: Global or Simultaneous?
Back to course
[Please upgrade your browser to play this video content]
Video Transcription
Hello, and this is Atul Verma, and it is my pleasure to give this talk on contact activation mapping for atrial fibrillation, global or simultaneous, as part of this session. And I want to thank the organizers of HRS for inviting me to give this talk. Here are all of my relevant disclosures. So we know that there have been multiple methods of global mapping for atrial fibrillation that have identified both focal and rotational activations during human persistent atrial fibrillation. And in this session, we have outlined a number of them, including non-contact mapping, contact-based mapping, sequential mapping, and body surface mapping. And in all of these cases, we've seen that identification of these phenomena can be challenging and that automated algorithms may assist operators to identify and characterize these phenomena. So I was asked to talk about contact activation mapping, and more specifically, the system that I've had a lot of familiarity with is the CardoFinder module. This is basically a module within the regular Cardo system that takes recorded electrical signals collected from multi-electrode catheters as inputs. It then generates the output data that is displayed as static or dynamic 4D maps, which can assist in the diagnosis of cardiac arrhythmias. And keep in mind that this is using unipolar signal analysis for activation mapping. So if we talk a little bit about the regions of interest, something that also makes the CardoFinder algorithm interesting is that it does have built-in automated software to automatically identify focal activation centers that are identified by performing morphology-based analysis on unipolar signals to detect QS patterns. And then rotational activation patterns are identified by tracking data of sequential activation gradients, and I'll get into that in a little bit more detail. So if we look a little bit deeper at the methodology of the system, it basically collects 30-second recordings of unipolar signals, either from a basket or multipolar catheter. And here are the filter settings that you should be aware of. The signals undergo processing to remove ventricular far-field detections, and then the system automatically annotates atrial signals. Activation maps are then displayed, which references each of the electrograms to all of the others, and only electrodes within 10 millimeters of the geometry are projected to avoid internal projections. And I think very importantly, this is the key point here, that unipolar signals can be reviewed and manually re-annotated, and there is no use of phase mapping. So unlike some of the other systems where you are very much dependent on the activation maps in order to determine whether something exists or not, here you can look at the activation map, but you can also look at the recordings, the raw recordings, to make yourself comfortable with whether there really is something of interest or not at that particular section. So this is just a graphical summary of how the system works, where basically you have an intracardiac atrial fibrillation electrogram. The algorithm detects far-field ventricular electrograms, and then basically subtracts those ventricular electrograms from the atrial signal, so you have a pure atrial signal. Then there is a calculated bipolar electrogram for each of the electrodes that you see. A bipolar electrogram window is then determined, and then using those bipolar windows on unipolar signals, the system is then able to perform a wavelet analysis to automatically annotate those wavelets, and you see that down at the bottom of your screen. And then to generate the color three-dimensional or four-dimensional map, basically there is a color gradient that exists between sequential signals, and you can see that the red here represents the leading edge of that activation, whereas the blue represents the trailing edge of that activation. So these are the kinds of images that you can see. This is a looped image, so here we see a rotational activation that is moving in this counterclockwise direction. Keep in mind that this activation is not permanent. This is a looped recording, but more importantly, you can look at the electrograms over here on the right to confirm to yourself that there is, in fact, this sequential activation gradient that is consistent with the rotation that we're looking at. This is an example of a focal activation that is originating from this point, and again here you can see the activation, unipolar signals, and you can confirm the QS pattern, which would be consistent with the activation. Now initially, the Cardo-Finder system was using basket catheters in order to map atrial fibrillation, but since has moved to sequential use of a multipolar mapping catheter, in this case the Pentarray catheter, where you can keep the Pentarray for at least 30 seconds in each position of the atrium. And then what these blue and green annotations represent is the results of the automated algorithm, which are basically telling you the locations of focal activations, which are represented in green, or rotational activations, which are represented in blue. So it becomes a color cue by which the operator can then target these rotational focal activations. Now we did some analysis of the automated region of interest algorithm. So again, just to give you some more detail on how it works, in order to identify a focal activation, it basically identifies QS patterns, and it needs to identify more than three early QS patterns within a two-second window in order to identify that as a focus. And for rotational activations, we have to identify a pan-systolic activation, which occupies at least 50% of the local cycle length, with a distance of less than 20 millimeters from the start to end point, and two or more such activations need to be identified within the 30-second period. Now, some of the mapping systems suggest that these focal or rotational activations occur throughout time, and are very, very stable temporally. However, we found that the majority of these activations were actually intermittent in nature, and were usually not lasting throughout the entire 30-second duration. And this is data from Richard Schilling's group in Britain, which was published in Jackie P. in 2018. And what they basically showed was that the majority of the rotational and focal activations tended to occur about two to three times within the 30-second recording, and a relative minority of them tended to occur more frequently than that over the course of the 30 seconds. Now, we did some of our own analysis, which was published in JCE of 2018. And what we found was that sequential mapping with a pentarray catheter was certainly feasible as long as you measured enough of a time duration at each position of the catheter. So here you can see that this is the percentage of rotational activation patterns or focal activation patterns that are identified. And you can see that if you only keep the catheter in a location for 1 or 5 or 10 seconds, you're only going to identify about half of all of the phenomena. But once you keep your recording catheter there between 15 and 20 seconds, you start to be able to identify nearly 100% of all of the rotational and focal activations. And what we found was that when we then process these activations through the automated algorithm, the algorithm performed very well with the sensitivity and specificity that was in the 85% to 95% range for focal and rotational activations. If we look at where the focal activations are located, here are drawings of the left atrium and the right atrium. You will see that the majority of these were occurring in both of the atria, with many of them occurring very close to the right atrial appendage, near the left atrial appendage as well, and quite a few of them located as well on the posterior wall of the left atrium. If we look at rotational activations, we saw a very similar distribution with about 60% near the right atrial appendage, a quarter of them near the left atrial appendage, but a relatively smaller proportion seen on the posterior wall, and most of them seen really around the appendageal regions. If we look at the average intensity of focal events seen over the 30 seconds, we see that it was about 15 of these events occurring over the 30 seconds, and if we look at the intensity of rotational events, we see that on average it was about 20 during the 30 seconds. But if we look at the frequency of the types of activations, by far focal activations were much more frequently seen than rotational activations, where focal activations were seen in nearly all of the recordings, whereas rotational activations were really only seen in a minority of the recordings. What happens when we did not detect a region of interest, so we did not see any focal or rotational activations? The most common reasons for not detecting them were either because we had extremely scarred atria, or very fractionated signals, or a combination of both. What's also interesting is that there seems to be an interrelationship between the focal and rotational activations. So if I talk you through this slide, most of the focal activations tended to occur without an associated rotational activation. So in other words, foci could exist without a rotation. However, only about half of the rotations occurred alone, with the vast majority of them only occurring with an associated focal activation. And interestingly, the majority of rotations tended to occur within 15 millimeters of a focal activation, which then raises the point, perhaps, are the foci actually driving the rotations? And it's kind of interesting, because when we look at some historical data, in this case, Waldo's data from Circulation 2016, and some more data from Fedorov's group in Circ-EP in 2016, what Waldo found was that the majority of activation patterns in human atrial fibrillation were actually focal drivers, as opposed to rotational drivers. And Fedorov also found that even when you have a micro-reentrant driver, they often will present as focal breakthroughs. So it makes sense, perhaps, that the foci are more predominant, and are perhaps driving the rotations. And furthermore, in our data, when we looked at patients who had acute termination of their atrial fibrillation versus those who did not, the acute termination was more significantly associated with ablation of focal drivers, and less associated with the ablation of rotation drivers. And if we look at some clinical observations, and again, this is the data from Schilling's group in JAK-EP 2018, they identified 26 drivers in 20 patients. 22 of these gave a clinical response, with 12 resulting in AF termination, and 10 resulting in AF cycle length prolongation greater than 30 milliseconds. And it's interesting that quite a high proportion of these were, in fact, focal activations. Now I've had the privilege of working not only with the CardoFinder system, but also with multiple other technologies, including the Acutis non-contact mapping system, the body surface mapping system from CardioInsight. And what I found is that there are actually far more similarities than differences in the phenomena that are being identified, which is certainly reassuring, because if every technology was showing something different, then I think we would have to be very concerned. I believe there may be a hierarchy of phenomena to ablate, maybe starting with the pulmonary veins, moving to the focal activations, and then perhaps the rotational activations. And there may be important interrelationships between the dynamic physiology as well as anatomical segments of the left and right atria. So I think we need to have a healthy skepticism, but I think there is enough interest in this area of mapping the dynamic substrate that we have included this as one of the arms for STAR-AF3, where we're going to be comparing PVI to PVI plus posterior wall isolation and elimination of drivers in our upcoming study. And with that, I would like to thank you very, very much for your attention and for the privilege of this presentation.
Video Summary
In this talk, Atul Verma discusses the contact activation mapping for atrial fibrillation using the CardoFinder module. This module takes recorded electrical signals and generates static or dynamic 4D maps to assist in diagnosing cardiac arrhythmias. The system uses unipolar signal analysis for activation mapping and has built-in software to identify focal and rotational activations. The CardoFinder algorithm performs well with sensitivity and specificity in the 85% to 95% range. The majority of focal and rotational activations were found to be intermittent and not lasting throughout the entire 30-second recording. The system uses the Pentarray catheter for sequential mapping, where focal activations were found in both atria, and rotational activations were mostly near the appendage regions. Foci were more predominant and often associated with rotations. Acute termination of atrial fibrillation was more significantly associated with the ablation of focal drivers rather than rotational drivers. Overall, contact activation mapping has similarities with other mapping technologies and may be a crucial component in the treatment of atrial fibrillation.
Keywords
contact activation mapping
atrial fibrillation
CardoFinder module
unipolar signal analysis
focal and rotational activations
Heart Rhythm Society
1325 G Street NW, Suite 500
Washington, DC 20005
P: 202-464-3400 F: 202-464-3401
E: questions@heartrhythm365.org
© Heart Rhythm Society
Privacy Policy
|
Cookie Declaration
|
Linking Policy
|
Patient Education Disclaimer
|
State Nonprofit Disclosures
|
FAQ
×
Please select your language
1
English