Skip main navigation

Ripple-AT Study

A Multicenter and Randomized Study Comparing 3D Mapping Techniques During Atrial Tachycardia Ablations
Originally published Arrhythmia and Electrophysiology. 2019;12:e007394



    Ripple mapping (RM) is an alternative approach to activation mapping of atrial tachycardia (AT) that avoids electrogram annotation. We tested whether RM is superior to conventional annotation based local activation time (LAT) mapping for AT diagnosis in a randomized and multicenter study.


    Patients with AT were randomized to either RM or LAT mapping using the CARTO3v4 CONFIDENSE system. Operators determined the diagnosis using the assigned 3D mapping arm alone, before being permitted a single confirmatory entrainment manuever if needed. A planned ablation lesion set was defined. The primary end point was AT termination with delivery of the planned ablation lesion set. The inability to terminate AT with this first lesion set, the use of more than one entrainment manuever, or the need to crossover to the other mapping arm was defined as failure to achieve the primary end point.


    One hundred five patients from 7 centers were recruited with 22 patients excluded due to premature AT termination, noninducibility or left atrial appendage thrombus. Eighty-three patients (pts; RM=42, LAT=41) completed mapping and ablation within the 2 groups of similar characteristics (RM versus LAT: prior ablation or cardiac surgery n=35 [83%] versus n=35 [85%], P=0.80). The primary end point occurred in 38/42 pts (90%) in the RM group and 29/41pts (71%) in the LAT group (P=0.045). This was achieved without any entrainment in 31/42 pts (74%) with RM and 18/41 pts (44%) with LAT (P=0.01). Of those patients who failed to achieve the primary end point, AT termination was achieved in 9/12 pts (75%) in the LAT group following crossover to RM with entrainment, but 0/4 pts (0%) in the RM group crossing over to LAT mapping with entrainment (P=0.04).


    RM is superior to LAT mapping on the CARTO3v4 CONFIDENSE system in guiding ablation to terminate AT with the first lesion set and with reduced entrainment to assist diagnosis.

    Clinical Trials Registration: Unique identifier: NCT02451995.


    • Three-dimensional activation mapping of atrial tachycardias is difficult in scarred atria, and a diagnosis is usually made in conjunction with entrainment.

    • Ripple mapping shows electrograms as moving bars on a voltage map and enables diagnosis without annotation or window-of-interest errors.


    • This is the first prospective, multi-center and randomized study comparing mapping techniques for atrial tachycardia.

    • Ripple mapping was superior to standard 3D activation mapping methods.

    • Ripple mapping also reduced the need for entrainment for making a diagnosis.


    The incidence of atrial tachycardia (AT) has increased with the rising numbers of atrial fibrillation (AF) ablation.1 Three-dimensional mapping and ablation for patients with symptomatic ATs is proven to be superior to pharmacological therapy and is the recommended first-line treatment.2

    The current gold standard approach to AT mapping relies on the annotation of local activation time (LAT) of each intracardiac electrogram collected within a prespecified window of interest (WOI) based on the tachycardia cycle length (TCL).3 This approach can be prone to error, especially in areas of low voltage related to prior ablation, surgery or myopathy where electrogram annotation can be challenging.4 Three-dimensional mapping systems continue to develop algorithms to overcome the challenges of LAT mapping in these areas without addressing the fundamental limitations related to annotation. Ripple mapping (RM) is now an established alternative approach to activation mapping on the CARTO3 CONFIDENSE platform (Biosense Webster, Inc) that does not require electrogram annotation or a WOI.5–7 Furthermore, as Ripple maps can be played over a color display of bipolar voltage, it can demonstrate how activation navigates through areas of low voltage.8 Recent nonrandomized studies have suggested that RM can improve diagnostic accuracy compared with standard LAT mapping approaches.8,9

    In this study, we prospectively randomized patients with AT to test whether RM is superior to LAT mapping on the CARTO3 version4 (v4) CONFIDENSE platform in diagnosing the mechanism of AT to guide the delivery of radiofrequency ablation.


    Study Design

    The Ripple-AT study was a multi-center, prospective, and randomized study comparing RM versus LAT guided AT ablation. Centers experienced in the 3D mapping of ATs and using the CARTO3v4 CONFIDENSE module (which includes CARTO Ripple Mapping as a module) volunteered for participation. All operators received formal training in the use of RM, and undertook at least 4 consecutive RM guided AT ablations before commencing the randomized study. All procedures were performed with informed consent, and institutional approval at each site was granted for this study. The data that support the findings of the study are available from the corresponding author on reasonable request.

    Patients were recruited from those referred for paroxysmal/persistent AT ablation. Subjects were excluded if the documented ECG was consistent with typical cavo-tricuspid isthmus dependent flutter. Patients were block randomized into an unblinded 1:1 mapping design of either RM or LAT mapping to guide ablation. Randomization was performed via a sealed envelope.

    A decapolar catheter was placed in the coronary sinus (CS), and a suitable CS reference was selected. Burst pacing down to 200ms from different CS poles was used to induce AT if in sinus rhythm at the start. The CS activation pattern in AT was used to decide on the first chamber to map. CARTO3v4 CONFIDENSE (Biosense, Inc) was used for mapping with a Lasso Nav (Biosense, Inc) or Pentaray Nav (Biosense, Inc) as per operator discretion. A point density to color the entire geometry with a color threshold of 5 mm was targeted. Criteria for including points on the map using CONFIDENSE Continuous Mapping included a cycle length stability within a 5% range of the TCL, an electrode position stability within 2 mm, an LAT stability filter at 3 milliseconds, and tissue proximity to the endocardial surface. While industry support was offered to assist with operating the CARTO system, they provided no assistance in interpreting the activation map to which each patient was randomized.

    Randomized to RM

    RM presents the entire electrogram as a white moving bar on the surface geometry with the height correlating to the electrogram voltage amplitude at that time point. Every bipolar electrogram from each point is aligned relative to the reference signal, and the Ripple map of all points is then played through time, with activation understood as the movement of bars traverses from one area to the next.

    The approach to RM undertaken by operators in this arm of the study is presented in Table 1. Details of how this approach was developed can be seen in previous published studies.7,8 Electrogram deflections above 0.07 mV were displayed as Ripple bars but could be adjusted down to 0.03 mV to review very low voltage activation. The bars were clipped above 0.30 mV to allow easier visualization. The map was played as a continuous loop spanning 2 TCLs. Electrogram data from the same tachycardia cycle is presented as a Ripple map by aligning the electrograms according to their corresponding reference electrogram from the same cycle.

    Table 1. Protocol for Atrial Tachycardia Diagnosis Using Ripple Mapping

    Ripple Mapping of an Atrial Tachycardia—Diagnostic Steps
    MappingMultipolar mapping—Lasso/PentaRay
    Colour threshold 5 mm: point density=no grey areas (usually >2000 points).
    Only “points projected” displayed (visualisation set-up)
    Annular points tagged and mitral/tricuspid annulus removed
    Ripple setupRipple preferences: Show bars above—0.07 mV; Clip bars above—0.30 mV
    Selected point viewer: 2 cycle lengths played
    Surface voltage thresholdingBipolar voltage map displayed—set empirically to 0.3–0.3 mV on custom settings
    Play Ripple map and reduce voltage limits in 0.05 mV steps (ie, 0.25–0.25 mV; continued down to 0.05–0.05 mV as required) until no Ripple wave-fronts in red areas (nonactivating tissue)
    Ripple bar wave-fronts only visible in purple areas
    Identifying mechanismStudy Ripple activation in small patches of geometry. Use design lines with arrowheads to mark these wave-fronts throughout the entire chamber
    Follow the arrows backwards to identify focal source or reentrant circuit.
    Reduce Show bars above to 0.03 mV—study activation in areas of interest, to locate (1) earliest bar of focal source or (2) narrowest isthmus of the re-entrant circuit.

    In this arm of the study, during the geometry and electro-anatomic map acquisition, the bipolar voltage map was displayed and the LAT map remained hidden to all lab staff. The voltage display was set empirically to 0.30 to 0.30 mV (such that tissue <0.30 mV was colored red, and tissue >0.30 mV was colored purple). The voltage color scale was manually reduced from 0.30 mV to identify the surface voltage threshold supporting wave-fronts of Ripple bars. In doing so, areas supporting Ripple wavefronts were displayed in purple, and the remaining nonconducting areas without Ripple bars in red (eg, 0.15–0.15 mV, where tissue below 0.15 mV appeared red, and tissue above 0.15 mV appeared purple). Every potential Ripple activation wavefront was studied to determine the AT mechanism. In complex cases, the design line tool dialogue box was activated, which allowed the operator to manually draw out the path of the visualized activation direction across the entire geometry chamber (as small white arrows), to plan the ablation approach.

    Randomized to LAT Mapping

    Conventional activation mapping involves the measurement of the LAT of each electrogram. This denotes the numerical difference in timing between a component of a sampled electrogram and a stable reference signal, the values of which are plotted on the 3D map according to a rainbow color bar that spans the interval of the mapped tachycardia. Annotation of LAT requires a timing WOI to ensure that electrograms from the same beat are compared. A WOI is defined by specifying a timing interval both before and after the reference point. The position of the reference point in relation to the window is arbitrary. The color red is used to represent the earliest measurement of LAT within the specified window and purple the latest. In a reentrant tachycardia, the concept of early and late is a misnomer, because any given area of activation will have other sites that are activated before or after that location, and the red area of earliest activation will be adjacent to a purple area of latest activation (ie, early meets late).

    The CARTO3v4 CONFIDENSE module offers several features to enhance LAT mapping. This includes Wavefront annotation, an automated algorithm that assigns timing annotation at the maximal negative distal unipolar derivative within the time-window spanned by the corresponding bipolar electrogram. In addition, there is an algorithm to enable an automated WOI recompute, which allows the WOI and the subsequent LAT of each projected point to be recalculated at any time during map acquisition. There is also a map consistency filter, which determines the consistency of each measured LAT point relative to its neighboring points, and highlights outlier points with LAT’s very different to its surrounding neighbors for review and deletion/reannotation as required.

    In this arm of the study, during the geometry and electro-anatomic map acquisition, the LAT and bipolar voltage maps were displayed and the Ripple map remained hidden to all lab staff. The WOI was set from the start, spanning between 90% and 100% of the TCL and either around the reference signal or the surface P wave.10 Operators were able to adjust the WOI as required using the automated recalculation function for the complete electroanatomic map. Automated annotation of activation timing was assigned using the CONFIDENSE Wavefront algorithm in all cases. In the absence of a sharp negative slope >0.03 mV in unipolar signal, a grey square was projected as the system could not automatically assign LAT. After all points were collected, the Map Consistency filter was used as required. The static LAT map was then visualized as a 2 color propagation map (red and blue). Additional manual LAT reannotation or deletion was allowed at sites where propagation remained unclear on the map. In a reentrant circuit, an early meets late algorithm could be applied to interpolate between early and late sights on the map considered reentrant.


    The AT diagnosis was considered macro-reentrant if it traveled continuously around the mapped atrial chamber (where head met tail), while small loop circuits were contained within a single plane (ie, anterior wall). Both focal/localized reentrant circuits appeared to emanate from a small focus, spreading radially away and often involving low voltage and fractionated electrograms at the break-out site.


    A diagnosis was made using the mapping arm assigned and a strategy for ablation was planned. Power controlled (25–35 W) radiofrequency energy was delivered (Stockert 70 RF generator, Biosense Webster) through an irrigated ablation catheter (ThermoCool SmartTouch, Biosense Webster) at the putative target for AT ablation.

    End Points

    This study protocol was applied only to the first map collected (ie, AT1). The primary end point of the study was AT termination with the first ablation set. The first ablation set encompassed all the ablation lesions delivered to target the AT based on the studied activation map. For example, if the mapping approach suggested LA roof dependency, the first ablation set would have been a roof line(s) and AT termination after this ablation set alone would achieve the primary end point. If AT1 changed to AT2 with ablation (defined as a sustained change in CS activation or cycle length), this was also considered to have met the primary end point.

    Cases where the tachycardia (1) terminated to sinus rhythm or (2) degenerated to AF/alternating ATs, before mapping was completed, were excluded from analysis. The objective of this study was to assess the diagnostic efficacy in the acute setting only, therefore no long-term data were collected.


    In this study protocol, operators were asked to make a diagnosis with RM or LAT alone, and entrainment was restricted to allow a fair comparison of the 2 mapping approaches in isolation. Thus, if the operator was confident of the AT diagnosis and ablation strategy from the 3D map, no entrainment was performed. A single entrainment manuever was allowed where the operator had some diagnostic uncertainty, to validate the diagnosis before ablation. For example, if the mapping approach suggested peri-mitral macro-reentry, entrainment from CS proximal and CS distal would have been permissible as a single confirmatory entrainment manuever. If the diagnosis remained uncertain, or if ablation failed to terminate AT, this was considered a failure to meet the primary end point of the study. Crossover to the other mapping arm of the study and additional entrainment mapping was open as per operator discretion.


    A prior nonrandomized study demonstrated a diagnostic yield using RM compared with LAT maps of 90% versus 65%.9 Assuming the same effect size, to detect a difference between arms with 80% power at the 5% 2-tailed significance level, this required at least 40 patients in each group. A sample size of at least 80 patients was targeted.

    Categorical variables were expressed as percentages. Continuous variables were expressed as mean±1 SD for parametric data and median (lower quartile—upper quartile) for non-normal data. Categorical data were analyzed using either a Fisher exact test or χ2 test where appropriate. Unpaired data were analyzed using a Student t test for parametric data and Mann Whitney U test for non-normal data. A 2-sided P value was determined where applicable, and a value of P≤0.05 was considered significant.


    Figure 1 summarises the study design and number of patients in each arm. A total of 105 patients were recruited from 7 centers (6 United Kingdom; 1 Portugal) participating in this study. Of these patients, 22 were excluded. This was due premature termination of AT while mapping (n=6), AT noninducibility (n=9), or degeneration to alternating ATs/AF before the delivery of ablation (n=4). Left atrial appendage thrombus on TOE seen at the start of the procedure (n=3) resulted in immediate procedural cessation.

    Figure 1.

    Figure 1. Study design and patient numbers. AT indicates atrial tachycardia; CTI, cavo-tricuspid isthmus; LAA, left atrial appendage; LAT, local activation time; RA, right atrium; and RM, ripple mapping.

    The remaining 83 patients completed mapping and AT ablation as per the assigned randomized arm, and these were used for subsequent analysis. This included 42 patients in the RM group and 41 patients in the LAT group.

    The baseline demographics, prior ablation history, and mapping details between the 2 groups were similar, and is summarized in Table 2. RM versus LAT: (65±9 years) versus (65±10 years) P=0.78; prior atrial ablation n=32 (76%) versus n=33 (80%) P=0.83—this included prior pulmonary vein isolation n=30 (71%) versus n=29 (71%); linear lesions/complex fractionated atrial electrogram ablation n=20 (48%) versus n=12 (29%) P=0.13; prior right atrial ablation only n=2 (5%) versus n=4 (10%) P=0.65. Patients with prior cardiac surgery without ablation included (RM versus LAT) n=3 (7%) versus n=2 (5%) P=0.66. Therefore, the total number of patients with potential iatrogenic scar causing AT (prior ablation or cardiac surgery) included RM n=35 (83%) versus LAT n=35 (85%) P=0.80.

    Table 2. Baseline Demographics and Mapping Details

    Baseline CharacteristicsAssigned GroupP Value
    No. of patients4241
    Prior atrial ablation or cardiac surgery (%)83851.00
    Age, y65±965±100.78
    Prior atrial ablation (%; RA or LA)76800.80
    Prior PVI (%)7171NA
    Prior LA substrate ablation (eg, lines/CFAE; %)48290.12
    Prior RA ablation only (%)5100.43
    Prior cardiac surgery without ablation (%)751.00
    AT cycle length/ms, median267*260*0.2
    Pentaray use (%)64630.93
    Collected Points, median2681*22190.44

    AT indicates atrial tachycardia; CFAE, complex fractionated atrial electrogram; LA, left atrium; LAT, local activation time; PVI, pulmonary vein isolation; RA, right atrium; and RM, Ripple mapping.

    *Denotes 3 missing values.

    †Denotes 4 missing values.

    The median AT cycle lengths was similar between groups (RM versus LAT: 267 ms [LQ 240, UQ 298] versus 260 ms [LQ 231, UQ 278], P=0.20). Multipolar catheters (Lasso and Pentaray) were used for mapping in almost all cases (a linear Smarttouch catheter was used in the noncoronary cusp in a focal AT) with similar median point densities between groups RM versus LAT: 2681points (LQ 1722, UQ 3647) versus 2219 points (LQ 1453, UQ 3232), P=0.44. The Pentaray was the most preferred mapping catheter RM versus LAT: n=27 (64%) versus n=26 (63%), P=0.93.

    Figure 2 summarizes the key end points of the study. The primary end point of termination of AT with the first ablation set occurred in 38/42 patients (pts; 90%) in the RM group and 29/41 (71%) in the LAT group (P=0.045). This was achieved with the 3D map alone, that is, without entrainment in 31/42 pts (74%) in the RM group and 18/41 pts (44%) in the LAT group (P=0.01). AT1 changed to AT2 with ablation in 6 pts in both groups, and the AT2 map confirmed a different mechanism and ablation target in all cases. One patient in each group failed to terminate with ablation where entrainment confirmed the same diagnosis as the map. Where the primary end point was not met (ie, those without a diagnosis based on the 3D mapping method, or where the first ablation set failed to terminate AT), AT termination was achieved in n=9/12 (75%) patients crossing from LAT to a combination of RM and entrainment mapping, and in n=0/4 (0%) patients crossing from RM to a combination of LAT and entrainment mapping (P=0.04).

    Figure 2.

    Figure 2. Results of acute ablation outcomes. AT indicates atrial tachycardia; and LAT, local activation time.

    Where the primary end point was met, left-sided ATs were predominantly identified in both groups (RM versus LAT: n=31/38 [82%] versus n=24/29 [83%], P=0.90). Macro/small loop reentrant mechanisms were more common in both groups (RM versus LAT: 27/38 [71%] versus 22/29 [76%], P=0.87; with the remaining diagnosed as focal/localized reentrant (RM versus LAT: 11/38 [29%] versus 7/29 [24%]).

    There were no procedural complications in either groups in this study. Figure 3 (and Movie I in the Data Supplement) illustrates a case of a patient randomized to RM. A complex LA circuit circumnavigating around an island of probable scar <0.15 mV was observed on the anterior wall, determined by the absence of Ripple wave-fronts through it. A second circuit was also observed around the mitral annulus (Movie I in the Data Supplement). Both these circuits were dependent on a narrow and slowly conducting isthmus between the inferior border of this island and the mitral annulus, and transecting this isthmus terminated tachycardia. This case demonstrates how RM is used to define putative scar on the map by functional assessment, and then exploiting the concurrent display of activation and scar to define the optimal ablation site.

    Figure 3.

    Figure 3. Randomized to Ripple mapping. Voltage thresholding defines the critical isthmus (See also Movie I in the Data Supplement). Patient with an idiopathic left atrial tachycardia (AT; 262 ms). A, Upper part—Anteroposterior (AP) bipolar voltage map set to 0.30 to 0.30 mV. Tissue with voltage <0.30 mV is displayed in red, and >0.30 mV in purple. A large area colored red was seen on the anterior wall. Wave-fronts of Ripple bars (yellow circle) were seen in areas of the voltage map colored red. B, The surface voltage limits were reduced until 0.15–0.15 mV (ie, red ≤0.15 mV; purple ≥0.15 mV)—here no Ripple bars were seen in areas colored red. At this voltage setting, an island of red tissue was seen on the anterior wall with Ripple bars circumnavigating clockwise around it (marked out by white arrows). C, Ripple markings were sampled around this island, and their corresponding electrograms (EGMs) spanned the atrial tachycardia cycle length (3 consecutive EGM cycles shown). Shown in the accompanying video is a second circuit traveling counter-clockwise around the mitral annulus. Both circuits were dependent on a narrow isthmus between the inferior border of this island and the mitral annulus. D, Transecting this critical isthmus with ablation terminated tachycardia.

    There were 4/42 cases which failed to meet the primary end point in the RM group. One case was peri-mitral (confirmed with entrainment) where MI block could not be achieved. One case terminated with ablation after further diagnostic mapping. One case targeted a septal source without effect. The last case involved extensive low voltage substrate/scar from multiple prior ablations where a roof substrate was targeted without effect.

    Figures 4 through 7 (and corresponding Movies II through IV in the Data Supplement) highlight cases of patients randomized to LAT mapping that did not meet the primary end point. The cases in Figures 4 and 5 are from patients with prior surgical atrial septal defect repair, and illustrate issues related to LAT color interpolation. In Figure 4, several early and late sites colored red and purple are seen on the final LAT display. As these early and late sites were in close proximity, the operator applied the early meets late tool to interpolate the colors between these sites and produce a uniform pattern. When propagation was played, the operator observed continuous rotation around scar on the anterolateral wall suggestive of small loop reentry. However, split potentials were identified along the circuit, implying wave-front collision, not observed on the propagation display. Given this uncertainty, the operator crossed over to RM. The Ripple map demonstrated that a line of conduction block prevented small loop reentry. Thus, the appearance of reentry was false, and a consequence of over-interpolation creating the impression of reentry.

    Figure 4.

    Figure 4. Randomized to local activation time (LAT) mapping. Interpolation creates the false impression of small loop reentry (See also Movie II in the Data Supplement). Patient with a prior surgical atrial septal defect repair mapped in atrial tachycardia (AT; 260 ms) in the right atrium. A, LAT map (modified AP) with the window of interest (WOI) set either side of the reference signal (−130 ms to +130 ms). More than one region of early and late was seen (labeled). B, The early meets late tool was applied (80%-default), and color interpolation between these early and late sites was filled (dark red). Scar was highlighted as areas with peak bipolar voltage amplitude ≤0.03 mV and colored grey. All the colors of the LAT spectrum were seen to progress around a patch of scar on the anterolateral wall suggestive of small loop reentry. However, split potentials were identified along the circuit, implying wave-front collision in the circuit. C, The Ripple map demonstrated that a line of conduction block prevented small loop reentry around this patch of scar, as evident by a line of double potentials along it (double yellow lines). In fact, Ripple bars rotated around this line of block, creating a larger loop reentry. The pattern of activation is depicted by white design lines, and can be viewed in the corresponding video. Entrainment supported this observation—the post-pacing interval was long within the interpolated false small loop circuit seen by the LAT map, and shortened moving superiorly into larger loop circuit seen with the Ripple map. RA indicates right atrium.

    In Figure 5, the full color-coded spectrum was observed within a small area collocating with an area of low voltage consistent with the lateral surgical cannulation site. The propagation map demonstrated wavefront turning around this site, and sampled electrograms were fractionated, leading the operator to again consider small loop reentry. This site was ablated without effect. The patient crossed over to RM, where no rotational activity was seen, rather splitting of activation on either side of this region of probable scar. Post-procedure, a band of false color interpolation spanning the full rainbow spectrum was appreciated on the LAT map between the apparent early and late sites on the map. This created the appearance of a slowly moving backward wave-front and this false appearance of wavefront turning on the propagation display. As RM does not require a WOI and does not interpolate, this error was avoided.

    Figure 5.

    Figure 5. Randomized to local activation time (LAT) mapping. Interpolation creates a false backward wavefront (See also Movie III in the Data Supplement). Another patient with prior surgical atrial septal defect repair was mapped in atrial tachycardia (AT; 270 ms) in the right atrium. A, The operator observed the full rainbow color spectrum around the lateral right atrium corresponding to (B) an area of low voltage <0.30 mV on the bipolar voltage map and considered this small loop reentry based on (C) the presence of long and fractionated electrograms (EGMs) within the circuit and the appearance of wavefront turning/curvature (although not a complete rotation) on the corresponding LAT propagation map (see video). However, ablation at this site (red circular VisiTag disks) was ineffective. D, There was no evidence of wavefront curvature on the Ripple map, rather splitting of wavefronts on either side of this region of scar. E, A band of false color interpolation spanning the full rainbow color spectrum was evident between the early and late sites on the LAT map (labeled) that created the appearance of a slowly moving backward wave-front and the impression of wave-front turning. CS indicates coronary sinus; IVC, inferior vena cava; SVC, superior vena cava; and WOI, window of interest.

    Figure 6 depicts an iatrogenic AT post extensive AF ablation where the LAT WOI had been set equally around the CS reference electrogram. The subsequent activation pattern appeared focal in origin from within the left atrial appendage. Given the absence of electrograms of interest, this diagnosis was uncertain, and entrainment revealed this site was outside the circuit. RM did not demonstrate focal activation from the left atrial appendage, rather activation breakout from probable scar near the posterior floor. Post-procedure, after multiple arbitrary post hoc adjustments of the LAT WOI, it revealed a similar activation pattern as seen with RM. This case highlights how setting the WOI is arbitrary and can be misleading in complex cases. As RM does not require a WOI, this limitation can be avoided.

    Figure 6.

    Figure 6. Randomized to local activation time (LAT) mapping. Varying LAT activation patterns when changing the window of interest (See also Movie IV in the Data Supplement). Patient with post atrial fibrillation (AF) ablation pulmonary vein isolation (PVI, Roof+Mitral isthmus lines, complex fractionated atrial electrogram [CFAE] ablation) atrial tachycardia (AT; 258 ms). Upper part: The window of interest (WOI) was set either side of the reference signal to span 90% of the tachycardia cycle length (TCL). The LAT map, seen in AP (A) and PA (B) depicts earliest focal activation (red) within the left atrial appendage (LAA). C, The corresponding voltage map (PA) demonstrates extensive substrate <0.1 mV (colored red). Unlike the LAT map, RM revealed a breakout of activation from probable scar near the posterior floor (yellow star), considered the exit of a region of slow conduction supporting a macro-reentrant circuit around the mitral annulus (white design lines). D, Adjustment of the LAT WOI to (−18 ms) to (+201 ms) revealed the same breakout activation pattern (red isochrone collocating with yellow star) as suggested by Ripple mapping (RM), but considered the mechanism focal rather than part of a macro-reentrant circuit. AT terminated in the mid-coronary sinus (purple star, snapshot catheter projection applied), presumably along with an epicardial connection of a mitral annular circuit. CS indicates coronary sinus; LLPV, left lower pulmonary vein; LUPV, left upper pulmonary vein; RLPV, right lower pulmonary vein; and RUPV, right upper pulmonary vein.

    Despite all attempts to optimize, the LAT map in Figure 7 was considered uninterpretable, with multiple early and late sites. This was mapped from a patient with prior mitral valvuloplasty and extensive low voltage atrial tissue. Electrograms sampled within the map were frequently long duration and multicomponent. Despite being in close proximity, sites have been labeled as early and late due to overlapping electrograms spanning a large portion of the set WOI. This case highlights the challenge of having to annotate a single LAT to represent multi-component fractionated electrograms. RM does not annotate, rather it presents all electrograms components in its entirety. Following crossover to RM, a breakout of activation from an extensive area of low voltage along the posterior roof was appreciated and a decision to perform a posterior box lesion resulted in AT termination.

    Figure 7.

    Figure 7. Randomized to local activation time (LAT) mapping: Uninterpretable activation pattern. Patient with post-surgical (mitral valvuloplasty) left atrial tachycardia (AT; 230 ms). The activation pattern (posteroanterior [PA]) was uninterpretable with multiple early and late sites (labeled). The electrograms (EGMs) from 2 immediately adjacent points are presented. These bipolar EGMs are low voltage and multicomponent, and span a large portion of the tachycardia cycle length (TCL). Their corresponding unipolar EGMs are also displayed. Despite being in immediate proximity, the system has annotated one (left) as early and the other (right) as late with respect to the set window of interest (WOI), based on the sharpest unipolar dV/dT within this long period of bipolar activation. Repeated occurrences of this event have resulted in this uninterpretable appearance. ATCL indicates atrial tachycardia cycle length; CS, coronary sinus; and Uni, unipolar.


    This study is the first prospective, multicenter, and randomized study comparing 3D activation mapping techniques. We show that operators experienced with LAT mapping on the CARTO3v4 CONFIDENSE platform had a higher rate of AT termination using RM, and achieved this with less reliance on entrainment support.

    The effectiveness of mapping tachycardia activation using LAT is proven. Published studies document ≈85% success rates in AT termination using this approach.1 LAT mapping has seen recent advances, including automated annotation to the maximum negative unipolar electrogram derivative within the period of bipolar activation, and its incorporation with high point density. However, most of these published studies have combined LAT with entrainment mapping such that the efficacy of LAT mapping in isolation remains unknown. Furthermore, these studies did not report whether ablation had been delivered at multiple incorrect sites before eventual AT termination. This is the first study to measure the efficacy of LAT mapping in isolation, without entrainment, and following the delivery of only the first ablation set. This study demonstrated acute AT termination with first lesion set in 71% with LAT mapping, and in only 44% without entrainment. The figures highlight 3 sources for error in relation to LAT mapping that likely explain this ≈30% failure rate, including (1) incorrect color interpolation; (2) WOI errors, and (3) mis-annotation.

    Interpolation algorithms assign the average activation time between mapped points to display an interpretable propagation pattern on the assumption that activation is uniform; however, these estimates of timing can be misleading, especially in areas of conduction delay or block, as seen in Figure 4. Backward wave-fronts are a specific interpolation error observed in macro-reentrant circuits, and caused the problems demonstrated in Figure 5. These occur at sites where early and late do not quite meet, and interpolation of colors between these apparent early and late sites occur, resulting in a slowly moving wave-front in reverse to the true direction of activation.

    A WOI is required to ensure that electrogram LATs from the same cycle are compared. However, the process of setting this window is arbitrary, with different color-coded activation patterns generated depending on the setting, as shown in Figure 6. While methods to standardize this approach around the surface P wave have been considered, they are only applicable in nonfocal mechanisms, which is not known at the start of the case.10 Furthermore, in diseased tissue with prolonged conduction times, no matter how the window is set, being limited to a single cycle can lead to very late activating sites being erroneously displayed as early in the WOI with respect to the next cycle of activation, resulting in more than one early site on the map.11 RM is the only contact-mapping system which currently reviews more than a single TCL. The CARTO3v4 LAT maps used in this study present their WOI as a color bar. A WOI color wheel has also been proposed to solve the challenges of setting a WOI. Rotation of this wheel has the equivalent effect of sliding the WOI without causing a full map recompute. In principle, this can avoid some errors related to the WOI as considered above. However, several studies have reported on a high prevalence of small pseudo-reentrant circuits from continuously rotating the wheel, some of which were misdiagnosed as localized/small loop reentry and inappropriately ablated.12,13

    Misannotation of LAT can lead to a complete change in the color-coded pattern. The advent of high point density acquisition with algorithms that filter out points with inconsistent timings in relation to neighboring LAT measurements has reduced annotation errors. However, these errors remain prevalent, particularly in areas of low voltage containing multicomponent electrograms as in Figure 7. Manual checking and reannotation of LAT when >2000 pts are collected is time-consuming and impractical during clinical procedures.

    While all 3D-mapping systems continue to develop algorithms to overcome these limitations to LAT mapping, RM offers a completely alternative activation mapping approach. RM presents activation information without the need for annotation of activation time or setting of a WOI, and does not interpolate between unmapped sites.5 Patients randomized to RM in this study achieved AT termination in 90% with the first ablation set (P=0.045).

    This superiority was partly attributable to avoiding these errors associated with LAT mapping. It was also consequent to a unique means of studying activation in areas of low voltage and scar. There remains no consensus on a voltage parameter to differentiate between active and nonconducting tissue (ie, either true scar from fibrosis or areas of functional block dependant on the wave-front direction and atrial rate) using endocardial mapping. LAT maps apply an arbitrary preset voltage threshold to display scar, and display areas below this threshold as grey tags to blank the color-coded map.14 With RM, displaying activation wave-fronts on a colored voltage map enables a novel approach to defining this voltage parameter that differentiates electrically active myocardium from nonconducting tissue (voltage thresholding).8 The example in Figure 3 is a case where only the common isthmus of the dual-loop tachycardia needed to be ablated. This was possible because of the simultaneous display of activating and nonconducting myocardium during tachycardia that helped determine the optimal site for ablation. This can be particularly helpful in peri-mitral tachycardias in which conventional mitral isthmus lines can be avoided.

    Entrainment enhances our electrophysiological understanding of the AT circuit before ablation, and this study does not advocate entrainment avoidance.15 However, entrainment does have limitations in areas of low voltage due to pacing latency, noncapture, and can cause degeneration to AF.16 In this study, patients randomized to RM underwent significantly less entrainment than those in the LAT group (P=0.01), as operators felt more confident to ablate the AT based on the Ripple map alone. Entrainment appears to be essential to LAT mapping to help overcome the core limitations of this approach, while with RM it was more supportive by confirming a diagnosis.

    RM looks very different to conventional LAT activation maps. There is a learning curve, and each operator (who had extensive experience in LAT mapping and entrainment) received 2 to 3 hours of formal training in RM on a workstation with example cases. This was followed by 4 consecutive clinical cases where operators made a diagnosis with RM first, with the improved diagnostic efficacy of RM already apparent at that stage.9 These operators did not need technical support from industry representatives to make these diagnoses. We would consider similar training to be essential for other operators aiming to replicate the results seen in this study.

    While this study was specific to mapping electrograms within the atria, RM is also suited to mapping electrograms within the ventricle. We and others have shown how RM can be used to follow late potentials through channels of slow conduction that might support reentry.17,18 RM to study VT and substrate during sinus rhythm has a very different workflow due to the main challenge being differentiating local and far-field electrograms.


    The objective of this study was to assess the diagnostic efficacy of each mapping technique in the acute setting, and does not consider the best approach to achieve long-term freedom from AT. The analysis of long-term outcomes of patients randomized to each mapping arm would require a different protocol that prohibited crossover to the other mapping arm during the entire case, and mandated postablation inducibility testing (which was performed at operator discretion in this protocol) and treatment of any subsequent ATs using the same mapping approach.

    These cases were all performed on the CARTO3v4 CONFIDENSE platform, and some of the findings might not apply to other mapping software.

    Approximately 20% of patients recruited were not included in the study analysis due to tachycardia termination before mapping was started/completed. We excluded these patients as the ATs were not stable and, therefore, our end point of acute termination during ablation would not have been robust. The best approach to achieve acute success and long-term benefit in this group of patients needs further study.


    This prospective, randomized, and multi-center study demonstrates that RM is superior to LAT mapping on the CARTO3v4 CONFIDENSE platform in achieving acute AT termination using the first delivered ablation set, with reduced reliance on entrainment to assist diagnosis.


    This abstract received the “Eric N. Prystowsky Fellows Clinical Research Award” at Heart Rhythm Society 2019. This abstract presentation also received a “Late Breaking Research Award” at Asia-Pacific Heart Rhythm Society 2018.


    The Data Supplement is available at

    Prapa Kanagaratnam, PhD, Department of Cardiology, Mary Stanford Wing, St Marys Hospital, Imperial College Healthcare NHS Trust, London W2 1NY, United Kingdom. Email


    • 1. Chae S, Oral H, Good E, Dey S, Wimmer A, Crawford T, Wells D, Sarrazin JF, Chalfoun N, Kuhne M, Fortino J, Huether E, Lemerand T, Pelosi F, Bogun F, Morady F, Chugh A. Atrial tachycardia after circumferential pulmonary vein ablation of atrial fibrillation: mechanistic insights, results of catheter ablation, and risk factors for recurrence.J Am Coll Cardiol. 2007; 50:1781–1787. doi: 10.1016/j.jacc.2007.07.044CrossrefMedlineGoogle Scholar
    • 2. Zhang XD, Gu J, Jiang WF, Zhao L, Zhou L, Wang YL, Liu YG, Liu X. Optimal rhythm-control strategy for recurrent atrial tachycardia after catheter ablation of persistent atrial fibrillation: a randomized clinical trial.Eur Heart J. 2014; 35:1327–1334. doi: 10.1093/eurheartj/ehu017CrossrefMedlineGoogle Scholar
    • 3. Jaïs P, Shah DC, Haïssaguerre M, Hocini M, Peng JT, Takahashi A, Garrigue S, Le Métayer P, Clémenty J. Mapping and ablation of left atrial flutters.Circulation. 2000; 101:2928–2934. doi: 10.1161/01.cir.101.25.2928LinkGoogle Scholar
    • 4. Del Carpio Munoz F, Buescher TL, Asirvatham SJ. Teaching points with 3-dimensional mapping of cardiac arrhythmias: teaching point 3: when early is not early.Circ Arrhythm Electrophysiol. 2011; 4:e11–e14. doi: 10.1161/CIRCEP.110.960344LinkGoogle Scholar
    • 5. Linton NW, Koa-Wing M, Francis DP, Kojodjojo P, Lim PB, Salukhe TV, Whinnett Z, Davies DW, Peters NS, O’Neill MD, Kanagaratnam P. Cardiac ripple mapping: a novel three-dimensional visualization method for use with electroanatomic mapping of cardiac arrhythmias.Heart Rhythm. 2009; 6:1754–1762. doi: 10.1016/j.hrthm.2009.08.038CrossrefMedlineGoogle Scholar
    • 6. Jamil-Copley S, Linton N, Koa-Wing M, Kojodjojo P, Lim PB, Malcolme-Lawes L, Whinnett Z, Wright I, Davies W, Peters N, Francis DP, Kanagaratnam P. Application of ripple mapping with an electroanatomic mapping system for diagnosis of atrial tachycardias.J Cardiovasc Electrophysiol. 2013; 24:1361–1369. doi: 10.1111/jce.12259CrossrefMedlineGoogle Scholar
    • 7. Koa-Wing M, Nakagawa H, Luther V, Jamil-Copley S, Linton N, Sandler B, Qureshi N, Peters NS, Davies DW, Francis DP, Jackman W, Kanagaratnam P. A diagnostic algorithm to optimize data collection and interpretation of ripple maps in atrial tachycardias.Int J Cardiol. 2015; 199:391–400. doi: 10.1016/j.ijcard.2015.07.017CrossrefMedlineGoogle Scholar
    • 8. Luther V, Linton NW, Koa-Wing M, Lim PB, Jamil-Copley S, Qureshi N, Ng FS, Hayat S, Whinnett Z, Davies DW, Peters NS, Kanagaratnam P. A prospective study of ripple mapping in atrial tachycardias: A novel approach to interpreting activation in low-voltage areas.Circ Arrhythm Electrophysiol. 2016; 9:e003582. doi: 10.1161/CIRCEP.115.003582LinkGoogle Scholar
    • 9. Luther V, Cortez-Dias N, Carpinteiro L, de Sousa J, Balasubramaniam R, Agarwal S, Farwell D, Sopher M, Babu G, Till R, Jones N, Tan S, Chow A, Lowe M, Lane J, Pappachan N, Linton N, Kanagaratnam P. Ripple mapping: Initial multicenter experience of an intuitive approach to overcoming the limitations of 3D activation mapping.J Cardiovasc Electrophysiol. 2017; 28:1285–1294. doi: 10.1111/jce.13308CrossrefMedlineGoogle Scholar
    • 10. De Ponti R, Verlato R, Bertaglia E, Del Greco M, Fusco A, Bottoni N, Drago F, Sciarra L, Ometto R, Mantovan R, Salerno-Uriarte JA. Treatment of macro-re-entrant atrial tachycardia based on electroanatomic mapping: identification and ablation of the mid-diastolic isthmus.Europace. 2007; 9:449–457. doi: 10.1093/europace/eum055CrossrefMedlineGoogle Scholar
    • 11. Ju W, Yang B, Chen H, Zhang F, Gu K, Yu J, Li M, Yang G, Cao K, Chen M. Mapping of focal atrial tachycardia with an uninterpretable activation map after extensive atrial ablation: tricks and tips.Circ Arrhythm Electrophysiol. 2014; 7:598–604. doi: 10.1161/CIRCEP.114.001508LinkGoogle Scholar
    • 12. Luther V, Sikkel M, Bennett N, Guerrero F, Leong K, Qureshi N, Ng FS, Hayat SA, Sohaib SM, Malcolme-Lawes L. Visualizing localized reentry with ultra-high density mapping in iatrogenic atrial tachycardia: beware pseudo-reentry.Circ Arrhythm Electrophysiol. 2017; 10:e004724.LinkGoogle Scholar
    • 13. Bradfield JS, Huang W, Tung R, Buch E, Okhovat JP, Fujimura O, Boyle NG, Gornbein J, Ellenbogen KA, Shivkumar K. Tissue voltage discordance during tachycardia versus sinus rhythm: implications for catheter ablation.Heart Rhythm. 2013; 10:800–804. doi: 10.1016/j.hrthm.2013.02.020CrossrefMedlineGoogle Scholar
    • 14. Kalman JM, VanHare GF, Olgin JE, Saxon LA, Stark SI, Lesh MD. Ablation of ‘incisional’ reentrant atrial tachycardia complicating surgery for congenital heart disease. Use of entrainment to define a critical isthmus of conduction.Circulation. 1996; 93:502–512. doi: 10.1161/01.cir.93.3.502LinkGoogle Scholar
    • 15. Pathik B, Lee G, Nalliah C, Joseph S, Morton JB, Sparks PB, Sanders P, Kistler PM, Kalman JM. Entrainment and high-density three-dimensional mapping in right atrial macroreentry provide critical complementary information: entrainment may unmask “visual reentry” as passive.Heart Rhythm. 2017; 14:1541–1549. doi: 10.1016/j.hrthm.2017.06.021CrossrefMedlineGoogle Scholar
    • 16. Vollmann D, Stevenson WG, Lüthje L, Sohns C, John RM, Zabel M, Michaud GF. Misleading long post-pacing interval after entrainment of typical atrial flutter from the cavotricuspid isthmus.J Am Coll Cardiol. 2012; 59:819–824. doi: 10.1016/j.jacc.2011.11.023CrossrefMedlineGoogle Scholar
    • 17. Luther V, Linton NW, Jamil-Copley S, Koa-Wing M, Lim PB, Qureshi N, Ng FS, Hayat S, Whinnett Z, Davies DW. A prospective study of ripple mapping the post-infarct ventricular scar to guide substrate ablation for ventricular tachycardia.Circ Arrhythm Electrophysiol. 2016; 9:e004072.LinkGoogle Scholar
    • 18. Xie S, Kubala M, Liang JJ, Yang J, Desjardins B, Santangeli P, van der Geest RJ, Schaller R, Riley M, Supple G, Frankel DS, Callans D, Pac EZ, Marchlinski F, Nazarian S. Utility of ripple mapping for identification of slow conduction channels during ventricular tachycardia ablation in the setting of arrhythmogenic right ventricular cardiomyopathy.J Cardiovasc Electrophysiol. 2019; 30:366–373. doi: 10.1111/jce.13819CrossrefMedlineGoogle Scholar


    eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

    Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.