Impact of filtering upon ventricular tachycardia identification by correlation waveform analysis

Area of Science:

• Biomedical Engineering
• Cardiology
• Signal Processing

Background:

• Signal analysis enhances ventricular tachycardia (VT) detection in implantable antitachycardia devices.
• Morphological analysis of VT can enable personalized therapies for patients with multiple VT types.
• Correlation waveform analysis (CWA) shows promise for distinguishing VT from sinus rhythm (SR).

Purpose of the Study:

• To evaluate the impact of digital filtering on CWA performance for VT detection.
• To determine optimal filter passbands for accurate VT identification in intraventricular recordings.

Main Methods:

• Bipolar intraventricular recordings of SR and 25 VTs from 18 patients were analyzed using CWA.
• Digital filtering was applied to recordings across various passbands (e.g., 1-500 Hz, 1-250 Hz, 1-100 Hz, 10-80 Hz).
• The separability of SR and VT at a 95% confidence level was assessed for each filter setting.

Main Results:

• Filtering with a 1-250 Hz passband yielded results equivalent to wideband analysis (≥92% SR/VT separability).
• Lower bandwidths (e.g., 1-100 Hz, 1-80 Hz) reduced discrimination, while specific ranges like 10-80 Hz improved it (88%).
• Typical pacemaker/defibrillator filter settings maintained high CWA performance, indicating robustness.

Conclusions:

• Digital filtering does not degrade overall VT detection accuracy when applied to CWA.
• Specific filter passbands can maintain or even enhance the reliability of CWA for differentiating VT from SR.
• CWA with optimized filtering is a viable method for improving VT detection in antitachycardia devices.