ECGI vs. FIRM: Direct Comparison, Phase/Waveform Mapping
Report by Dr. Steve S. Ryan, PhD
In a further discussion of the ECGI mapping and ablation system, Dr. Phillip Cuculich of the Washington University School of Medicine in St. Louis, MO gave a presentation entitled “Advances in and Limitations of Noninvasive Mapping of AF.” (For a detailed description and discussion of the ECGI system, see 2013 BAFS: Non-Invasive Electrocardiographic Imaging ECG (ECG).
Background: ECGI stands for Non-Invasive Electrocardiographic Imaging used at Yoram Rudy’s lab at Washington University in St. Louis to understand the mechanisms of heart rhythm disease. A similar system called Electrocardiographic Mapping (ECM/ecVUE) uses similar technology, but has been developed and tested for clinical use in Europe (http://www.cardioinsight.com) and has different goals. Each group works independently and has different ways to seek solutions.
The software used by the ECGI system to produce data and images from the multi-channel ECG mapping and CT scan is called CADIS.
POTENTIAL BENEFITS OF ECGI
Dr. Cuculich began by describing the potential benefits of ECGI:
• Save time: Locate the arrhythmia in a single beat
• Better Preparation: Understand and plan for the arrhythmia before an ablation
• Avoid Frustration: Map and ablate unstable, transient or complex arrhythmias
• Research Platform for Discovery: Identify and describe the mechanisms of arrhythmias
A-FIB PATTERNS OR SIGNALS AS REVEALED BY ECGI
After imaging patients with ECGI, A-Fib patterns are a combination of mechanisms:
• In simple A-Fib, Dr. Cuculich showed movies of left pulmonary vein focal sites with 1 to 2 wavelets and a left-to-right activation pattern
• In complex (Long-standing Persistent) A-Fib, he showed movies of four or more simultaneous wavelets, a high degree of wavelet curvature, and frequent wave breaks (no focal sites). The patterns tend to repeat and follow a preferred path.
FIRM AND ECGI COMPARISON
|Inside the heart||Outside the heart (body surface mapping)|
|Up to 64 contact electrodes to produce up to 64 electrograms||1000 reconstructed electrodes|
|QRST subtracted||No signal subtraction|
|70% rotor, 30% focal||Multiple wave fronts (1-4), 15% rotor|
|Stable beat-to-beat||Transient focal activity, transient rotational circuits|
(In the ECGI imaging/mapping system the number of points on the heart is changeable. But they have found 1000 to be a good number for reliable, detailed analysis.)
Dr. Cuculich compared ECGI data to recent invasive epicardial (inside-the-heart) mapping and body surface mapping (called “Phase Lock”). The data showed significant agreement between the imaging systems. Also, ECGI compares favorably to surgical maze mapping data.1
But compared to FIRM, the most common patterns of A-Fib Dr. Cuculich found were multiple wavelets, with pulmonary vein and non-pulmonary vein focal sites. Rotor activity was seen rarely.2
NO STANDARD DEFINITION OF “ROTOR”
There is no standard definition of a rotor. In Dr. Cuculich’s studies he used 2 rotations at the same spot as a “rotor.” This is perhaps why he found less rotors than in the FIRM system and in the CardioInsight system as described by Dr. Jais where they found 80% rotors. See: BAFS 2014 Jais, ECGi & Circular Catheter
Another major difference in ECGI and FIRM is that ECGI uses wavelet analysis (activation of the wavefront), while FIRM and CardioInsight uses phase mapping to describe the behavior of the arrhythmia. The main point of Dr. Cuculich’s presentation is that one must be very careful when applying phase techniques, as it can introduce rotor behavior into the imaging map. Dr. Cuculich’s group is studying whether this rotor behavior may be a true cause for the maintenance of A-Fib or just an artifact.
ECGI—TOO MANY ELECTRODES?
In a conversation with the author, Dr. Cuculich brought up comments that perhaps ECGI/ECM uses too many electrodes to see stable rotors, that perhaps panoramic imaging with fewer electrodes could improve the identification of rotors. (ECGI has a much larger number of electrode points in the heart [usually 1000] compared to FIRM [64 max].) To test this hypothesis, he analyzed A-Fib using 64 spaced electrodes in each atrium vs. standard ECGI. It turned out that fewer electrodes did not help to visualize rotors.
PHASE MAPPING, WAVEFRONTS, WAVELET TRANSFORMATION, ACTIVATION PATTERNS
Dr. Cuculich introduced new concepts in the use of ECGI (at least to this author)—phase mapping and the importance of wavefronts or wavelet transformation in A-Fib signals. “ECGI uses wavelet transform looking at pure activation time.” He asked, “how does…phase mapping affect the result?” He related phase mapping to the CONFIRM concept of “phase lock” where a simple 12-lead ECG analysis can classify A-Fib mechanistically.3
Doctors (and we patients) are still struggling to understand what phase mapping and wavelet transformation actually mean. Dr. Cuculich’s studies of phase mapping techniques (Hilbert transform) in A-Fib show that phase mapping highlights and accentuates the curvature of a wavefront and thus indicates a rotor is present. According to Dr. Cuculich, phase mapping is highly dependent on the chosen cycle length. He concluded that “while published ECGI data used wavelet transform to identify activation patterns, phase mapping techniques (when performed carefully and correctly) may offer additive information.”
We’re grateful to Dr. Cuculich for his comparison of ECGI and FIRM which helps us understand both imaging system better. But it’s definitely disturbing that both systems vary so greatly. Why does the FIRM system find 70% rotors and ECGI only 15%? Why are FIRM’s A-Fib signals stable and ECGI’s transient? Why does the FIRM system not focus on wavefronts and wavelet transformation?
One way to resolve these discrepancies would be to use a standard Lasso mapping catheter to meticulously map every potential A-Fib-producing spot in an animal or human with Long-standing Persistent A-Fib (where one would expect to find multiple A-Fib producing spots in the heart). Then immediately use both the FIRM and ECGI system to map the same heart and compare the results.
Perhaps the single biggest new discovery in human A-Fib mapping is rotors. But there’s considerable debate about their definition and behavior. Dr. Cuculich found that rotors are relatively rare (15%), whereas the FIRM and CardioInsight studies indicate that 70-80% of A-Fib drivers are rotors.
Dr, Cuculich introduced new concepts, insights and vocabulary to our understanding of A-Fib, (some of which I’m still having trouble wrapping my head around). Are wavefronts and wavelet transformation important in themselves or are they part of the development of rotors? Phase mapping and wavelet transformation applied to A-Fib is a major innovation that may lead to a better understanding of how A-Fib signals activate in the heart. Besides making mapping and ablating A-Fib easier and more effective, ECGI with its detailed, high resolution capabilities may give us new insights into A-Fib.
- Lee, G. et al. Epicardial wave mapping in human long-lasting persistent atrial fibrillation: transient rotational circuits, complex wavefronts, and disorganized activity. European Heart Journal (2104) 35, 86-97. Last accessed May 13, 2013, URL:http://www.ncbi.nlm.nih.gov/pubmed/23935092 doi:10.1093/eurheartj/eht267↵
- Cuculich, PS et al. Noninvasive characterization of epicardial activation in humans with diverse atrial fibrillation patterns. Circulation. 2010 Oct 5; 122(14): 1365-72. Last accessed May 13, 2013, URL: http://tinyurl.com/okp4229↵
- Non-invasive identification of stable rotors and focal sources for human atrial fibrillation: mechanistic classification of atrial fibrillation from the electrocardiogram, Europace. February 28, 2013. Last accessed May 13, 2013, URL: http://tinyurl.com/njt9zd7; doi:10.1093/europace/eut038↵