Yoko Ito1, Keisuke Shiga2, Kentaro Yoshida3, Kuniomi Ogata4, Akihiko Kandori4, Takeshi Inaba1, Yoko Nakazawa1, Yukio Sekiguchi1, Hiroshi Tada5, Kensuke Sekihara2, Kazutaka Aonuma1. 1. Cardiovascular Division, Institute of Clinical Medicine, University of Tsukuba, Tsukuba, Japan. 2. Department of Systems Design & Engineering, Tokyo Metropolitan University, Hachioji, Japan. 3. Cardiovascular Division, Institute of Clinical Medicine, University of Tsukuba, Tsukuba, Japan. Electronic address: kentaroyo@nifty.com. 4. Central Research Laboratory, Hitachi Ltd, Kokubunji, Japan. 5. Cardiovascular Division, Institute of Clinical Medicine, University of Fukui, Yoshida-gun, Japan.
Abstract
BACKGROUND: Although several reports address characteristic 12-lead electrocardiographic findings of outflow tract ventricular arrhythmias (OT-VAs), the accuracy of electrocardiogram-based algorithms to predict the OT-VA origin is sometimes limited. OBJECTIVE: This study aimed to develop a magnetocardiography (MCG)-based algorithm using a novel adaptive spatial filter to differentiate between VAs originating from the aortic sinus cusp (ASC-VAs) and those originating from the right ventricular outflow tract (RVOT-VAs). METHODS: This study comprised 51 patients with an OT-VA as the target of catheter ablation. An algorithm was developed by correlating MCG findings with the successful ablation site. The arrhythmias were classified as RVOT-VAs or ASC-VAs. Three parameters were obtained from 3-dimensional MCG imaging: depth of the origin of the OT-VA in the anteroposterior direction; distance between the earliest atrial activation site, that is, sinus node, and the origin of the OT-VA; and orientation of the arrhythmia propagation at the QRS peak. The distance was indexed to the patient's body surface area (in mm/m2). RESULTS: Origins of ASC-VAs were significantly deeper (81 ± 6 mm/m(2) vs. 68 ± 8 mm/m(2); P < .01) and farther from the sinus node (55 ± 9 mm/m2 vs. 41 ± 9 mm/m(2); P < .01) than those of RVOT-VAs. ASC-VA propagation had a tendency toward rightward axis. Receiver operating characteristic analyses determined that the depth of the origin was the most powerful predictor, with a sensitivity of 90% and a specificity of 73% (area under the curve = 0.90; P < .01). Discriminant analysis combining all 3 parameters revealed the accuracy of the localization to be 94%. CONCLUSION: This MCG-based algorithm appeared to precisely discriminate ASC-VAs from RVOT-VAs. Further investigation is required to validate the clinical value of this technique.
BACKGROUND: Although several reports address characteristic 12-lead electrocardiographic findings of outflow tract ventricular arrhythmias (OT-VAs), the accuracy of electrocardiogram-based algorithms to predict the OT-VA origin is sometimes limited. OBJECTIVE: This study aimed to develop a magnetocardiography (MCG)-based algorithm using a novel adaptive spatial filter to differentiate between VAs originating from the aortic sinus cusp (ASC-VAs) and those originating from the right ventricular outflow tract (RVOT-VAs). METHODS: This study comprised 51 patients with an OT-VA as the target of catheter ablation. An algorithm was developed by correlating MCG findings with the successful ablation site. The arrhythmias were classified as RVOT-VAs or ASC-VAs. Three parameters were obtained from 3-dimensional MCG imaging: depth of the origin of the OT-VA in the anteroposterior direction; distance between the earliest atrial activation site, that is, sinus node, and the origin of the OT-VA; and orientation of the arrhythmia propagation at the QRS peak. The distance was indexed to the patient's body surface area (in mm/m2). RESULTS: Origins of ASC-VAs were significantly deeper (81 ± 6 mm/m(2) vs. 68 ± 8 mm/m(2); P < .01) and farther from the sinus node (55 ± 9 mm/m2 vs. 41 ± 9 mm/m(2); P < .01) than those of RVOT-VAs. ASC-VA propagation had a tendency toward rightward axis. Receiver operating characteristic analyses determined that the depth of the origin was the most powerful predictor, with a sensitivity of 90% and a specificity of 73% (area under the curve = 0.90; P < .01). Discriminant analysis combining all 3 parameters revealed the accuracy of the localization to be 94%. CONCLUSION: This MCG-based algorithm appeared to precisely discriminate ASC-VAs from RVOT-VAs. Further investigation is required to validate the clinical value of this technique.