Abstract 2167: Spectral Analysis of Near-Field Electrograms of Ventricular Fibrillation in Implantable Cardioverter-Defibrillators
Studies have shown that 11–12% of patients with implantable cardioverter-defibrillators (ICDs) receive inappropriate shocks. Etiologies of shocks include lead malfunction and electromagnetic interference (EMI). Limited work has been done to distinguish inappropriate sensing due to lead malfunction or EMI from ventricular fibrillation (VF). A retrospective analysis was performed of 28 patients with implanted ICDs undergoing elective defibrillation threshold (DFT) testing. Fifty-six induced VF events and 56 baseline rhythm (BR) events were analyzed. Thirty-eight cautery and pin-header artifact events from 18 patients were also analyzed. These events were retrieved from the storage diskette and analyzed off-line using Matlab software. A power spectral density analysis using a Fourier transform was performed on a 2 second segment of the near-field data for VF, BR, and artifact. The fundamental frequency was located between 3.5–7.5 Hz. A ratio of the sum of the energy of the first five harmonics to the energy of the entire spectrum (Energy Ratio) and a ratio of the root mean square of the first five harmonics to the entire spectrum (RMS Ratio) were calculated for VF, BR, and artifact. Statistical comparison was made to VF using nonparametric methods. Data are expressed as median (25th, 75th percentiles). VF Energy and RMS Ratios were significantly different from BR and Artifact Energy and RMS Ratios. The range (min, max) for the Energy Ratios of VF (0.5530, 0.9920) and Artifact(0.0590, 0.5680) showed overlap. No overlap was found between the RMS Ratios of VF (0.9020, 1.0000), BR (0.0840, 0.8530) and Artifact (0.0460, 0.7740). Signal obtained during VF has unique spectral characteristics that may be used to distinguish VF from EMI or lead malfunction.