Correlation-based pattern recognition for implantable defibrillators
1Department of Electrical Engineering, Stanford University, California, USA.
Insights
New implantable devices can now detect and correct cardiac arrhythmias. A novel, computationally efficient system accurately distinguishes supraventricular tachycardia (SVT) from ventricular tachycardia (VT) using correlation-based morphology assessment.
Area of Science:
- Biomedical Engineering
- Cardiology
- Signal Processing
Background:
- Cardiac arrhythmias cause significant mortality, with limited treatment options historically.
- Implantable devices offer in vivo arrhythmia management but face power and computational constraints.
- Current heart rate-based classification algorithms have high error rates, particularly in distinguishing SVT from VT.
Purpose of the Study:
- To develop a computationally efficient arrhythmia classification architecture for implantable devices.
- To improve the accuracy of distinguishing between supraventricular tachycardia (SVT) and ventricular tachycardia (VT).
- To address the limitations of current rate-based algorithms and computationally intensive morphology assessment.
Main Methods:
- A novel correlation-based morphology assessment architecture was developed.
- Individual heartbeats are classified by comparing signal vectors to prestored templates.
- A series of beat classifications inform the overall rhythm assessment.
- The system leverages new pattern recognition techniques.
Main Results:
- The proposed architecture achieved excellent accuracy in discriminating between SVT and VT.
- The correlation-based approach is computationally efficient, suitable for implantable devices.
- The system overcomes limitations of traditional morphology assessment.
Conclusions:
- The developed computationally-efficient, correlation-based architecture enables accurate arrhythmia classification in implantable devices.
- This approach offers a significant improvement over existing rate-based methods for distinguishing SVT from VT.
- The findings pave the way for more effective in vivo management of cardiac arrhythmias.
Abstract:
An estimated 300,000 Americans die each year from cardiac arrhythmias. Historically, drug therapy or surgery were the only treatment options available for patients suffering from arrhythmias. Recently, implantable arrhythmia management devices have been developed. These devices allow abnormal cardiac rhythms to be sensed and corrected in vivo. Proper arrhythmia classification is critical to selecting the appropriate therapeutic intervention. The classification problem is made more challenging by the power/computation constraints imposed by the short battery life of implantable devices. Current devices utilize heart rate-based classification algorithms. Although easy to implement, rate-based approaches have unacceptably high error rates in distinguishing supraventricular tachycardia (SVT) from ventricular tachycardia (VT). Conventional morphology assessment techniques used in ECG analysis often require too much computation to be practical for implantable devices. In this paper, a computationally-efficient, arrhythmia classification architecture using correlation-based morphology assessment is presented. The architecture classifies individuals heart beats by assessing similarity between an incoming cardiac signal vector and a series of prestored class templates. A series of these beat classifications are used to make an overall rhythm assessment. The system makes use of several new results in the field of pattern recognition. The resulting system achieved excellent accuracy in discriminating SVT and VT.
More Related Videos
Related Concept Videos
01:30Pulse rhythm
Conversely, an irregular pulse pattern is termed dysrhythmia, stemming from disruptions in cardiac muscle...
01:25Correlation between ECG and Cardiac Cycle
A cardiac action potential originates in the SA node and spreads throughout the atria and the AV node in approximately 0.03 seconds. This results in the P wave in an ECG and triggers atrial contraction. The action potential is then briefly slowed at the AV node, allowing the atria to contract and fill the ventricles with blood before...