Image integration using intracardiac ultrasound to guide catheter ablation of atrial fibrillation.

BACKGROUND: Three-dimensional (3D) reconstruction of the left atrium (LA) can be performed using real-time intracardiac echocardiography (ICE) to facilitate image integration during atrial fibrillation (AF) ablation. Current users of this technology generally image the LA indirectly from the right atrium (RA). OBJECTIVE: The purpose of this study was to assess the feasibility and accuracy of image integration with placement of the ICE catheter directly in the LA to visualize the LA. METHODS: Thirty consecutive patients undergoing an AF ablation with the CARTO-Sound system were enrolled. A 10-Fr phased-array ICE catheter was used to obtain two-dimensional echocardiographic images of the LA; in 15 patients the ICE probe was placed in the LA, and in 15 patients it was placed only in the RA. Sequential images were obtained and merged with a preacquired computed tomography/magnetic resonance image. The accuracy of image integration was assessed by the value of the average image integration error after surface registration. RESULTS: Thirty patients (60% paroxysmal AF, LA size = 42 +/- 7 mm, ejection fraction = 62% +/- 10%) were studied. There was no difference in the time required to create the LA anatomic map and perform image integration with imaging from the LA versus the RA (22 +/- 22 vs. 24 +/- 16 minutes; P = .8). The number of ultrasound contours obtained was also similar (LA = 26 +/- 17 vs. RA = 24 +/- 16; P = .7). The average integration error was less with direct LA imaging (LA = 1.83 +/- 0.32 vs. RA = 2.52 +/- 0.58 mm; P = .0004). CONCLUSION: Direct LA imaging with ICE is feasible and results in improved LA visualization and image integration.

Effect of presenting rhythm on image integration to direct catheter ablation of atrial fibrillation.

INTRODUCTION: Magnetic resonance (MR) imaging of the left atrium (LA) can be integrated with electroanatomic mapping systems to guide catheter ablation of atrial fibrillation (AF). The usefulness of this technique is dependent on the accuracy of image integration. OBJECTIVE: The aim of this study is to determine the effect of heart rhythm at the time of pre-procedure MR imaging and heart rhythm at the time of ablation on integration error. METHODS: Fifty-two consecutive patients who underwent catheter ablation for AF were included. All patients underwent MR imaging of LA and pulmonary veins and image integration with real-time electroanatomic mapping. The rhythm at the time of MR imaging and on the day of ablation was recorded. CARTO-Merge software (Biosense-Webster) was used to calculate the average accuracy of integration of electroanatomic points with MR-derived reconstructions. RESULTS: There was no significant difference in integration error between patients who were in AF at the time of their MR vs. those who were in sinus rhythm at the time of their MR (1.76 +/- 0.26 vs. 1.88 +/- 0.31 mm, p = 0.15). There was also no significant difference in integration error between patients who were in concordant vs. discordant rhythms at the time of MR vs. day of ablation (1.81 +/- 0.23 vs. 1.89 +/- 0.32 mm, p = 0.40). There was a trend toward less integration error between patients who were in AF on the day of ablation vs. those in sinus rhythm (1.74 +/- 0.26 vs. 1.89 +/- 0.31 mm, p = 0.07). CONCLUSIONS: Image integration can be performed to direct catheter ablation of AF regardless of the rhythm at the time of imaging and ablation.

Pain and anatomical locations of radiofrequency ablation as predictors of esophageal temperature rise during pulmonary vein isolation.

INTRODUCTION: Esophageal temperature rise (ETR) during ablation inside left atrium has been reported as a marker for esophageal thermal injury. We sought to investigate the possible relationships between chest pain and ETR during radiofrequency (RF) ablation, and ETR and locations of RF application, in patients undergoing pulmonary vein (PV) isolation under moderate sedation. METHODS AND RESULTS: We analyzed anatomical locations of each RF application and its association with esophageal temperature and presence/absence of pain. Data from 40 consecutive patients (mean age: 56 +/- 10 years) were analyzed. There were a total of 4,071 RF applications resulting in 291 episodes of pain (7.1%) and 223 ETRs (5.5%). Thirty-five patients (87.5%) experienced at least one pain episode and 32 (80.0%) had at least one ETR. While 77.4% of posterior wall applications that caused pain also corresponded to an ETR (P < 0.0001), only 0.8% of pain-free posterior wall applications were associated with ETRs (P < 0.0001). The sensitivity and specificity of pain during ablation for ETR were 94% and 98%, respectively. No ETRs were observed during anterior wall applications. ETRs occurred more frequently during ablation on the left (86.1%) versus the right (13.9%), and in inferior (70.4%) versus superior (29.6%) segments. CONCLUSION: In patients undergoing PV isolation, ETR was encountered when ablating in the posterior left atrium with the distribution left > right and inferior > superior. Pain during ablation was associated with ETR, and lack of pain was strongly associated with absence of ETR. Pain during RF ablation may thus serve as a predictor of esophageal heating and potential injury.

Factors affecting error in integration of electroanatomic mapping with CT and MR imaging during catheter ablation of atrial fibrillation.

OBJECTIVE: Integration of 3-D electroanatomic mapping with Computed Tomographic (CT) and Magnetic Resonance (MR) imaging is gaining acceptance to facilitate catheter ablation of atrial fibrillation. This is critically dependent on accurate integration of electroanatomic maps with CT or MR images. We sought to examine the effect of patient- and technique-related factors on integration accuracy of electroanatomic mapping with CT and MR imaging of the left atrium. MATERIALS AND METHODS: Sixty-one patients undergoing catheter-based atrial fibrillation (AF) ablation procedures were included. All patients underwent cardiac CT (n = 11) or MR (n = 50) imaging, and image integration with real-time electroanatomic mapping of the aorta and left atrium (LA). CARTO-Merge software (Biosense-Webster) was used to calculate the overall average accuracy of integration of electroanatomic points with the CT and MR-derived reconstructions of the LA and aorta. RESULTS: There was a significant correlation between LA size assessed by electroanatomic mapping (112 +/- 31 ml) and average integration error (1.9 +/- 0.6 mm) (r = 0.46, p = 0.0003). There was also greater integration error for patients with LA volume >/= 110 ml (n = 31) versus < 110 ml (n = 30) (p = 0.004). In contrast, there was no significant association between average integration error and paroxysmal versus persistent AF, left ventricular ejection fraction, days from imaging to electroanatomic mapping, or images derived from CT versus MR. CONCLUSIONS: Patients with larger LA volume may be prone to greater error during integration of electroanatomic mapping with CT and MR imaging. Strategies to reduce integration error may therefore be especially useful in patients with large LA volume.