Author + information
- Received October 16, 2017
- Revision received February 12, 2018
- Accepted February 22, 2018
- Published online July 16, 2018.
- David A. Igel, PhDa,∗ (, )
- Jon F. Urban, PhDa,
- James P. Kent, BSa,
- Bernard Lim, MD, PhDb,
- K.L. Venkatachalam, MDc,
- Samuel J. Asirvatham, MDd and
- Daniel C. Sigg, MD, PhDa
- aFocusStart LLC, Minneapolis, Minnesota
- bHeart and Vascular Program, Baystate Medical Center, Springfield, Massachusetts
- cCardiovascular Diseases, Department of Medicine, Mayo Clinic, Jacksonville, Florida
- dDivision of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota
- ↵∗Address for correspondence:
Dr. David A. Igel, FocusStart, LLC, 807 Broadway Street Northeast, Suite 148, Minneapolis, Minnesota 55413.
Objectives This study investigated whether delivering negative charge to catheter tips reduces thromboembolism during catheter ablation.
Background Radiofrequency (RF) ablation prevents atrial fibrillation that can cause stroke or death. However, ablation itself can cause stroke (2%) or silent ischemia (2% to 41%), possibly via particulate debris that embolizes after coagulum adherence to catheter surfaces. Coagulum formation on RF catheters can be prevented by applying negative charge, but it is unknown if charge reduces peripheral thromboembolism.
Methods Paired (Charge ON vs. OFF) endocardial RF ablations were performed in 9 canines using nonirrigated RF catheters. Continuous negative charge was delivered via −100 μA of DC current applied to ablation catheter electrodes. Intracardiac echocardiography was used to navigate the catheter and to monitor coagulum formation. In a subset of 5 canines, microemboli flowing through polyester tubing between the femoral artery and vein (extracorporeal loop) were monitored with bubble counters and inline filter fabric. After each ablation, catheter-tip coagulum and blood particles deposited on the filters were quantified using photography and imaging software (ImageJ, U.S. National Institutes of Health, Bethesda, Maryland).
Results Negative charge significantly decreased the extracorporeal loop median filter area covered by particles (n = 19 pairs) by 10.2 mm2 (p = 0.03), and decreased median filter particles by 349 (p = 0.03). Negative charge also decreased the percentage of the catheter tip surface area covered by coagulum (n = 39 pairs) by 7.2% (p = 0.03).
Conclusions Negative charge delivery to ablation catheter tips during RF ablation can reduce particulate embolization material in an extracorporeal loop, and potentially reduce thromboembolic risk associated with RF ablation.
Radiofrequency (RF) catheter ablation is a common therapy to prevent a cardiac arrhythmia (atrial fibrillation) that can cause stroke or death. Although successful ablation may prevent stroke, the ablation procedure itself, paradoxically, can cause stroke in 0.6% (1) to 2% (2,3) of cases. Even if stroke symptoms are not detected post-ablation, studies have suggested that cognitive dysfunction may arise in 13% to 20% of cases (4), and that these deficits may be associated with “silent” ischemic lesions in the brain (5). Magnetic resonance imaging (MRI) studies have reported incidences of “silent” ischemic brain lesions after ablation between 2% and 41% (6). Thromboembolic events may arise from coagulum formation on RF ablation catheters, and/or other catheters introduced into the left atrium, during the ablation procedures.
It has been previously demonstrated in bench and pre-clinical studies that coagulum formation on cardiac RF ablation electrodes can be prevented by applying a small negative charge (7,8). The hypothesized mechanism of action was that the negatively charged electrode surface repelled the negatively charged fibrinogen molecules, and consequently prevented fibrinogen adherence to the electrode. The effect was dose-dependent in that as the amplitude of electrical current was increased, the thickness of clots formed at the catheter tip decreased to a threshold of 100 μA of negative DC current, above which coagulum formation was prevented.
Reducing coagulum formation during RF ablation may reduce the risk of thromboembolic events (e.g., stroke, other cerebrovascular or peripheral vascular thromboembolic events), and potentially enable physicians to use higher RF energies, thereby increasing the efficacy and safety of cardiac RF ablation procedures. We sought to determine whether a reduction of catheter tip coagulum is associated with a reduction in thromboembolic events. The effect of negative charge on systemic embolization was quantified by monitoring embolic material flowing through an extracorporeal loop. Our hypothesis was that negative charge applied to nonirrigated catheters during RF ablation reduces embolic microparticles.
The experimental protocol was approved by the Institutional Animal Care and Use Committee at American Preclinical Services, LLC (Minneapolis, Minnesota).
Negative charge device
A standalone device that delivers negative charge via −100 μA of DC current to ablation catheter tips was developed (8). Briefly, the charge-application circuit was connected between the electrode tip of the ablation catheter and a ground patch electrode, in parallel with the RF generator. The circuit was powered by 9-V batteries, and had a knob to adjust the amount of charge applied to the ablation electrodes. The circuit design, in which its internal resistance is much larger than the resistance of its target load, is a simple means to convert a voltage source (battery) to a current source. The current applied to the electrode/tissue interface results in ionic charge movement in the blood and tissue. Typical variations in the electrode/tissue interface resistance have very little effect on the amount of charge being delivered to the tissue. The high internal resistance of the charge-application circuit in parallel with the ablation generator also limited shunting of the applied ablation energy away from the cardiac tissue. During experiments, continuous negative charge was delivered using −100 μA of DC current; a similar amount of DC current was effective at preventing coagulum formation in previous studies (8).
The effects of negative charge on thromboembolism were assessed via two series of animals: in Series 1, emboli were evaluated via qualitative and semiquantitative assessments; in Series 2, emboli were quantified using an extracorporeal loop (9). For both series, canines (mongrel, >28 kg) were anesthetized using acepromazine and propofol, intubated, and subsequently maintained on 1% to 3% isoflurane during mechanical ventilation. An 11-F sheath was placed in the left femoral artery for blood pressure monitoring. Vital signs were monitored throughout the procedure including heart rate, blood pressure, temperature, SpO2, and activated clotting time (ACT). An 11-F sheath was placed in the right external jugular vein for intracardiac echocardiography (ICE) access.
Both intracardiac and epicardial echocardiography were available to optimize near-field visualization of catheter tip and tissue for real-time coagulum monitoring. ICE was performed via 8- or 10-F AcuNav ICE catheters utilizing the Siemens Acuson Sequoia C512 system (Munich, Germany). Epicardial echocardiography was performed via an echocardiography probe placed on the left atrial epicardium via minithoracotomy performed in the left fourth intercostal space.
Transseptal puncture was performed using a transseptal needle (BRK, St. Jude Medical, St. Paul, Minnesota) under ICE and fluoroscopic guidance; after confirmation of tenting in the fossa ovalis, the transseptal needle was advanced through the septum. Subsequently, an 8.5-F deflectable sheath (Agilis NxT, St. Jude Medical, St. Paul, Minnesota) was inserted into the left atrium. An unfractionated heparin bolus was administered after left heart access was performed. The ACT was measured every 30 min, and additional boluses were given to maintain an ACT of >300 s. Supplemental fluids (saline and hetastarch) were administered to maintain blood pressure.
The extracorporeal blood loop was constructed between a femoral artery and vein (9). A 15-F cannula was placed into the right femoral artery and a 17-F cannula was placed into the left femoral vein (both Medtronic Bio-Medicus, Dublin, Ireland). The 15-F arterial and 17-F venous cannulae were connected via a loop of polymeric perfusion tubing (3/8-inch inner diameter, NovoSci PN SFB10, Conroe, Texas). An in-line filter housing (Pall Inc. #1119, Port Washington, New York) was fitted with a replaceable 73-μm polyester filtration fabric (Sefar-Petex #07-73-40; Sefar AG, Heiden, Switzerland) and placed in the extracorporeal circuit. Two microbubble detectors (BC100, GAMPT Ultrasonic Solutions, Merseburg, Germany) were positioned proximal and distal to the filter.
Thromboembolic event monitoring
During ablation, clot or embolic material production (gaseous and/or particulate) was monitored via the ICE catheter or epicardial probe. After each ablation, the size (length) of any measurable coagulum attached to the catheter and/or proximal tissue was measured before the ablation catheter was removed. In addition, a semiquantitative thrombus score was reported by an echocardiographer as 0 (none), 1 (trace or minimal), 2 (medium), or 3 (large) in a blinded fashion.
In series 2 animals, embolic material flowing through the extracorporeal loop circulation was monitored in addition to the catheter-tip coagulum measurements performed in the series 1 animals. Particulate and gaseous embolizations were measured quantitatively with the extracorporeal loop bubble detectors. The number and volume of microbubbles flowing through the extracorporeal loop were continuously measured during ablations using a PC-based software system (GAMPT Ultrasonic Solutions, Merseburg, Germany). The number of small particulate emboli deposited on the blood filters and the filter area covered with particles were assessed after each ablation.
RF ablation was applied via Blazer II nonirrigated catheters (5-mm tip) connected to a Maestro 4000 generator (Boston Scientific, Marlborough, Massachusetts). Test ablations were performed in the right atrium to verify instrument function prior to performing ablations in the left atrium and left ventricle. Ablations were applied for 120 s unless ablation was stopped automatically by the generator. Ablation energies were programed to elicit coagulum formation on the catheter tip (8). Power, temperature, and impedance were monitored continuously on a Bard EP recording system (Boston Scientific, Marlborough, Massachusetts), along with surface ECG and ablation catheter electrograms.
Between 2 and 4 ablation catheters were used during each experiment. Ablations were paired so that those with charge were performed in temporal proximity to ablations without charge at similar programmed settings and at similar locations in the heart. Charge application order (ON-OFF vs. OFF-ON) was randomized to ensure catheters contributed data to both the charge-ON and charge-OFF groups. Surgeons and echocardiographers were blinded to the ablation regimen. Contact force was not measured directly, but location and tissue contact were verified intraoperatively via ultrasound.
After each ablation, coagulum on the catheter tip was scored and measured via ICE (if dimensions could be discerned) before the ablation catheter was removed. After careful removal of the ablation catheter from the animal, catheter tips were photographed for subsequent image analysis, then cleaned via ultrasonic cleaner in enzymatic solution. (The ultrasonic cleaning procedure was verified by assessing scanning electronic microscopic images from a subset of catheters.) After catheter removal, the extracorporeal loop blood filter fabric was removed from its housing, photographed for image analysis, and subsequently stored in formalin.
At the end of the experiment, triphenyl tetrazolium chloride was administered systemically to delineate lesions. The animal was then euthanized using Euthasol (Virbac AH, Inc.). Heart, lungs, brain, and kidneys were removed and inspected for ablation lesions (heart) or ischemic lesions; the brain and kidneys were sectioned into 5-mm-thick slices for inspection.
Analysis of catheter and filter images
The amount of coagulum coverage on catheter tips was quantified by planimetry digitally assisted by the ImageJ software program (10). Images were calibrated via a millimeter ruler that was placed next to the catheter when each picture was taken. Image measurements included the percentage coverage of the catheter tip, amount of coagulum extending outside of the catheter border (an estimate of coagulum volume), and thickness of coagulum measured at the thickest point.
The number of particles deposited on the blood filters was also quantified via planimetry digitally assisted by ImageJ. Image measurements included the number of filter particles, particle sizes, and total filter area covered by particles. A subset of 4 representative filters were stained, sectioned, and assessed by a board-certified pathologist to determine the nature of material comprising the filter particles.
Analyses were performed via the R statistical package (11) as a randomized block design, where treatment (charge ON) and control (charge OFF) were applied in matched ablation pairs. Data were assessed for normality using Shapiro-Wilks tests and found to be non-normal, except for bubble detector data. Due to non-normality, nonparametric statistics were utilized, rather than analysis of variance. Rank transformations were performed, and Wilcoxon rank sum tests were performed to calculate p values. Medians were calculated using the Hodges-Lehmann estimator. The values of continuous variables are expressed as medians and confidence intervals (CIs), unless otherwise noted. To reduce the risk that multiple comparisons led to significant differences by random chance, a Benjamini-Hochberg method (with a false discovery rate of 0.05) was used to modify the p values. A value of p < 0.05 was considered significant.
Ablations (n = 120) were performed in 9 canines; 78 of these ablations were applied as pairs (39 ablations with charge, and 39 ablations without charge), with each pair of ablations applied at similar programmed settings and locations. There were 4 series 1 canines that contributed 20 ablation pairs, and 5 series 2 canines (in which extracorporeal loop data were collected) that contributed 19 ablation pairs.
Ablations that were performed at typical temperature and power settings elicited little to no coagulum, so ablations were performed above the recommended settings. However, median power delivered when charge was OFF (26.5 W [95% CI: 23 to 29 W]) was similar to that when charge was ON (26.5 W [95% CI: 26.5 to 29 W]). At these settings, there was no notable difference in steam pops (3 during charge-OFF ablation, and 2 during charge-ON ablation).
Figure 1 illustrates filter contents and catheter-adhered char that resulted from an ablation pair. Analysis of the charge OFF catheter tip (Figure 1A) indicated that 35% of the tip was covered with coagulum. Particle analysis of the blood filter (Figure 1C) resulted in a count of 916 particles that covered 24.6 mm2 of the filter. The median size of particles in the filter was 0.018 mm2 with a range from 0.01 to 0.376 mm2 (filter fabric pore size was approximately 0.005 mm2). The charge-ON ablation resulted in 13% of the catheter tip being covered with coagulum, and the filter contained 193 particles that covered 5.7 mm2 of the filter. The median particle size was 0.019 mm2.
Histology performed on filter samples, conducted by a pathologist blinded to the experimental protocols, identified abundant fibrin admixed with scattered inflammatory cells, erythrocytes, and platelets. In multiple sections, clumps of myocytes were also observed. The smallest visible particles typically consisted of fibrin with scattered inflammatory cells and macrophages.
Effect of charge on embolic material (series 2 animals)
For each ablation pair, estimates of embolic material collected after ablations with charge OFF were subtracted from estimates after charge-ON ablations (charge-ON n − charge-OFF n); thus, positive differences represent more emboli with charge-ON, and negative numbers represent more emboli with charge OFF.
Ablation without charge resulted in a median number of filter particles of 651 (95% CI: 346 to 1,014), which covered an area of 19.4 mm2 (95% CI: 8.9 to 34 mm2) (Table 1). Charge decreased the total filter surface area covered by particulate debris in 13 of 19 ablation pairs, with a median decrease of 10.2 mm2 (Figure 2A) (p = 0.03). Charge similarly decreased the number of filter particles, with a median decrease of 349 particles (Figure 2B) (p = 0.03). Charge application showed trends of decreased bubble volumes and bubble counts, although the effects did not reach significance (Figure 2).
Effect of charge on catheter coagulum (series 1 and 2 animals)
The coagulum coverage of catheter tips was evaluated in all animals (series 1 and 2; 39 ablation pairs). Ablation without charge resulted in a median catheter tip coverage of 27% (95% CI: 20% to 32%) (Table 2). Charge application significantly decreased catheter tip coverage on 24 of 39 pairs, with a median decrease of 7.2% (p = 0.03). Charge effect on catheter-tip coagulum was also observed via ICE in that both thrombus score and thrombus length decreased with charge (Table 2) (p = 0.03). The area of outside of the coagulum and maximum coagulum thickness, however, were not affected by charge.
The median impedance with charge ON was 16 Ω [95% CI: −61 to −4 Ω] lower than that with charge OFF (p = 0.03) (Table 2). The decrease in impedance with charge is consistent with the observed decrease in catheter coverage with charge, as the formation of catheter-tip coagulum and char can be associated with an impedance rise.
Effect of charge on arrhythmia
Premature contractions and tachycardias arising in the atria or the ventricles were identified from the electrograms recorded on the Bard system during ablations. The number of these arrhythmias during ablation with charge OFF (n = 30) was approximately equal to that with charge ON (n = 33).
Ablation lesions were identified in all hearts. A majority of atrial lesions were transmural, and many exhibited granular char material embedded in trabeculae. Lesion size was not assessed in these experiments; lesion size differences between ablations with and without charge were assessed previously and were found to be equivalent (8).
Peripheral organs were evaluated to determine whether the particulate emboli sampled in the extracorporeal loop were associated with tissue damage. Because both charge-ON and -OFF ablations were performed in each canine, these necropsy observations were not intended to identify a charge effect on tissue damage, but rather to determine whether damage is likely if similar ablations are performed in a survival animal study. Brain and kidneys were inspected from 7 of 9 canines. No ischemic damage was visible in brain tissue, although brain lesions smaller than the spatial resolution of our dissection (5-mm slices) may have been present. Multiple infarcts were observed in kidneys from 2 of the 7 canines. In both cases, multiple lesions were identified in the right kidney.
In this study, we extend the research of the negative charge effect on catheter tip coagulum (8) by evaluating negative charge effect on particulate embolization. We found that applying −100 μA of DC current to the tips of nonirrigated catheters during ablation reduced particulate material in an extracorporeal loop (Figure 2).
Lim et al. (8) found that coagulum coverage of catheter tips exhibited an all-or-nothing response to −100 μA of current application during RF ablation, with catheter tip coverage decreasing from 91% without charge to 0% with charge. We found that −100 μA of charge decreased catheter-adhered coagulum from a median of 27% to 17% (median paired difference decrease of 7.2%) (Table 2). The effect of charge on catheter char may have been less pronounced in our experiments compared with that of Lim et al. (8) because of multiple procedural differences, including that our ablation parameters may have been more aggressive. Lim et al. (8) reported ablation temperatures ranging from 65oC to 95oC when utilizing catheters that were similar to ours in technology and dimension (4-mm or 5-mm nonirrigated). Our median ablation parameters were 29 W and 95oC. In agreement with prior studies (12–15), we found little coagulum formation below catheter tip temperatures of 80oC during ablation. We titrated to more aggressive ablation parameters to better test the effect of negative charge on coagulum and particulate embolization. It is possible that negative charge is less effective at higher temperatures. It is also possible that the higher temperatures of our ablations elicited greater damage to endothelial surfaces, thus predisposing ablation lesions to emit endocardial debris and to activate a thrombin cascade (1). However, therapeutic doses of heparin were administered to limit the risk of thrombin-dependent coagulation cascade.
Many additional factors affect the temperature and power during cardiac ablation, such as blood flow and catheter contact pressure, which may have also contributed to the difference between our results and that of Lim et al. (8). Also, our techniques for measuring coagulum differed from that of Lim et al. (8); they measured coagulum via ICE and by image analysis of scanning electronic microscopic images, whereas we applied image analysis to digital photographic images of the catheter tips.
Our techniques for monitoring embolic material in an extracorporeal loop were similar to those used by Haines et al. (9) and Takami et al. (16). However, our bubble volumes and filter particles were more numerous. In Haines et al. (9), the bubble volumes during irrigated catheter ablations in swine (0.4 nl) were notably smaller than our volumes (2.26 μl with charge off) (Table 1). Takami et al. (16) reported high microbubble volumes only after tissue pops (15.6 μl), and observed far smaller volumes when tissue pops were absent (maximum of 0.062 μl).
Our filter analysis identified at least 125 microparticles in each of our filters. The Haines et al. (9) analyses noted particles in all filters associated with overlapping PVAC electrodes, although they observed particles in only 3 of 9 filters associated with irrigated-tip catheters. Takami et al. (16) identified a median of 5 microparticles in 44 of 126 filters, and no microparticles in the remaining filters. However, the total number of filters examined by Takami et al. (16) included filters after sheath manipulations (in absence of ablation). Our results may have differed because most of their ablations were performed using irrigated catheters, whereas our ablations were performed using nonirrigated catheters and at high temperatures. Our quantitative image analyses utilizing the ImageJ image-processing program may have been more sensitive to particle counts compared with their gross histopathological analyses.
Our study resulted in kidney lesions in 2 animals, and no cerebral lesions were observed. Haines et al. (9) reported a similar incidence of renal embolization and cerebral lesions, whereas Takami et al. (16) observed no embolic lesions. The low rate of ischemic damage suggests that a survival study may not be justified unless more extensive ablations are applied.
Preliminary experiments on irrigated catheters
Although the objective of this work was to further the work of Lim et al. (8) by evaluating the effect of charge on particulate embolization during nonirrigated catheter ablation, it has been of interest (Michaud et al. ) to evaluate effects during irrigated catheter ablation. Paired (charge ON vs. OFF) endocardial RF ablations were performed in 5 canines using irrigated RF catheters. Catheter-attached coagulum was monitored via ICE, and embolic material was assessed from extracorporeal loop blood filters in a subset of canines. Ablation energies (30 to 50 W) were applied for 60 to 120 s and at irrigation rates between 2 and 30 ml/min. Ablation pairs in which tissue pops occurred were excluded from analysis (to distinguish the effect of charge on embolic material from the effect of tissue pops on embolic material). Catheter tip coagulum was scored via ICE in 13 ablation pairs as 0 (none), 1 (trace or minimal), 2 (medium), or 3 (large). Negative charge decreased coagulum score in 8 of the 13 pairs (p < 0.05). No coagulum was identified on the remaining 5 pairs (Figure 3). Of the 8 ablation pairs in which extracorporeal loop data were collected, ablation without charge resulted in a similar filter particle count (p = NS) as that with charge (399 [95% CI: 201 to 651] charge OFF vs. 325 [95% CI: 248 to 490] charge ON), and a similar area of filter coverage (11.0 mm2 [95% CI: 6.0 to 14.4 mm2] charge OFF versus 8.5 mm2 [95% CI: 6.5 to 13.6 mm2] charge ON). These data suggest that negative charge may prevent clot formation on irrigated-tip catheters during RF ablation. Additional extracorporeal loop experiments are needed to identify an effect on particulate embolization.
Blood flow variation within cardiac chambers and catheter contact force are known to be significant variables in ablation. Contact force was not monitored directly, but location and contact variations between ablation pairs were minimized by keeping surgeon operator and approximate ablation location consistent between ablation pairs.
We performed ablations at higher temperatures that those commonly utilized, to create a positive control against which to compare the effect of charge application. One of the potential implications of this work is that a technology that prevents embolism may facilitate safe ablations at settings more aggressive than those utilized today. Similar experiments at lower temperatures are warranted, although are likely to require a substantially larger number of animals to distinguish a charge effect, because lower temperatures are likely to result in fewer thromboembolic particles.
Because multiple ablations with and without charge were performed in the same cardiac chambers in the same animals, comparison of the effects of charge on tissue lesions and other ablation sequelae were not possible. It was noted previously (8) that charge had minimal effect on lesion size. An evaluation of granular char deposits resulting from charged versus noncharged ablations would be an interesting addition to future experiments.
MRI assessments of the brain (or other organs) were not performed; a subsequent study may compare MRI images of animals in which charge-ON ablations were performed to animals with charge-OFF ablations. However, the current study was designed to compare multiple thromboembolic estimates arising from charge-ON and -OFF ablations within the same animal.
We applied the same amount of negative current as that reported by Lim et al. (8), who identified −100 μA as an optimal charge to inhibit coagulum formation in a dose response experiment. Results may have differed if charge amounts were varied; however, we observed no incremental benefit to higher charge in preliminary experiments.
Negative charge delivery to catheter tips during RF ablation can reduce embolization material in an extracorporeal loop, and potentially reduce thromboembolic risk.
COMPETENCY IN MEDICAL KNOWLEDGE: Future tools and procedures utilized by cardiac electrophysiologists may incorporate surface-charge technology on their blood contacting surfaces to complement standard anticoagulation techniques (e.g., heparinization) to prevent thromboembolic complications.
TRANSLATIONAL OUTLOOK: This work represents a first translational step to determine whether charge technology limits thromboembolic events. The next steps to further development include activities associated with regulatory clearance, such as safety testing.
The authors graciously thank Amy Olson for performing the ImageJ analyses.
Research reported in this publication was supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under Award Number R44HL127758. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. FocusStart also received material support from Boston Scientific. Drs. Igel, Urban, Kent, and Sigg are employees of FocusStart and own equity pertaining to the charge device. Dr. Igel owns equity in FocusStart. Drs. Lim, Venkatachalam, and Asirvatham own intellectual property rights and equity pertaining to the charge technology. Dr. Venkatachalam has served as a consultant for BioSig Technologies. Dr. Asirvatham has received consulting fees from FocusStart, LLC.
All authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Clinical Electrophysiology author instructions page.
- Abbreviations and Acronyms
- activated clotting time
- direct current
- intracardiac echocardiography
- Received October 16, 2017.
- Revision received February 12, 2018.
- Accepted February 22, 2018.
- 2018 American College of Cardiology Foundation
- Cappato R.,
- Calkins H.,
- Chen S.A.,
- et al.
- Medi C.,
- Evered L.,
- Silbert B.,
- et al.
- Forleo G.,
- Della Rocca D.G.,
- Lavalle C.,
- et al.
- Lim B.,
- Venkatachalam K.L.,
- Henz B.D.,
- Johnson S.B.,
- Jahangir A.,
- Asirvatham S.J.
- Haines D.E.,
- Stewart M.T.,
- Dahlberg S.,
- et al.
- R Core Team
- Yokoyama K.,
- Nakagawa H.,
- Wittkampf F.H.,
- Pitha J.V.,
- Lazzara R.,
- Jackman W.M.
- Petersen H.H.,
- Chen X.,
- Pietersen A.,
- Svendsen J.H.,
- Haunso S.
- Takami M.,
- Lehmann H.I.,
- Parker K.D.,
- Welker K.M.,
- Johnson S.B.,
- Packer D.L.
- Michaud G.F.,
- Kumar S.