@inproceedings{RoeschHeinkeKleimenhagenetal.2011,
  author    = {Lena R{\"o}sch and Matthias Heinke and Frank Kleimenhagen and Christoph Melzer and Rudiger Fink and Frank Weber and Bruno Ismer},
  title     = {Can the accuracy of intracardiac electrogram based CRT optimization be improved?},
  series = {Biomedical Engineering / Biomedizinische Technik},
  volume    = {56},
  number    = {S1},
  publisher = {Walter de Gruyter},
  address   = {Berlin, Boston},
  issn      = {0013-5585 (Print)},
  doi       = {10.1515/bmt.2011.856},
  pages     = {6},
  year      = {2011},
  abstract  = {Introduction: To simplify AV delay (AVD) optimization in cardiac resynchronization therapy (CRT), we reported that the hemodynamically optimal AVD for VDD and DDD mode CRT pacing can be approximated by individually measuring implant-related interatrial conduction intervals (IACT) in oesophageal electrogram (LAE) and adding about 50ms. The programmer-based St Jude QuickOpt algorithm is utilizing this finding. By automatically measuring IACT in VDD operation, it predicts the sensed AVD by adding either 30ms or 60ms. Paced AVD is strictly 50ms longer than sensed AVD. As consequence of those variations, several studies identified distinct inaccuracies of QuickOpt. Therefore, we aimed to seek for better approaches to automate AVD optimization. Methods: In a study of 35 heart failure patients (27m, 8f, age: 67±8y) with Insync III Marquis CRT-D systems we recorded telemetric electrograms between left ventricular electrode and superior vena cava shock coil (LVtip/SVC = LVCE) simultaneously with LAE. By LVCE we measured intervals As-Pe in VDD and Ap-Pe in DDD operation between right atrial sense-event (As) or atrial stimulus (Ap), resp., and end of the atrial activity (Pe). As-Pe and Ap-Pe were compared with As-LA an Ap-LA in LAE, respectively. Results: End of the left atrial activity in LVCE could clearly be recognized in 35/35 patients in VDD and 29/35 patients in DDD operation. We found mean intervals As-LA of 40.2±24.5ms and Ap-LA of 124.3±20.6ms. As-Pe was 94.8±24.1ms and Ap-Pe was 181.1±17.8ms. Analyzing the sums of As-LA + 50ms with duration of As-Pe and Ap-LA + 50ms with duration of Ap-Pe, the differences were 4.7±9.2ms and 4.2±8.6ms, resp., only. Thus, hemodynamically optimal timing of the ventricular stimulus can be triggered by automatically detecting Pe in LVCE. Conclusion: Based on minimal deviations between LAE and LVCE approach, we proposed companies to utilize the LVCE in order to automate individual AVD optimization in CRT pacing.},
  language  = {en}
}