Automatic detection of ECG cable interchange by analyzing both morphology and interlead relations

Area of Science:

• Biomedical Engineering
• Cardiology
• Signal Processing

Background:

• Electrocardiogram (ECG) cable interchange can lead to misdiagnoses.
• High specificity is crucial for ECG cable interchange detection algorithms due to low prevalence.
• Accurate detection of ECG cable interchange is vital for reliable cardiac diagnostics.

Purpose of the Study:

• To propose and evaluate an improved algorithm for automatic detection and classification of ECG cable interchange.
• To enhance the accuracy of ECG interpretation by minimizing errors from cable misconnections.
• To develop a robust system for identifying specific types of ECG cable errors.

Main Methods:

• Algorithm development incorporating ECG morphology and redundancy information.
• Utilized features like QRS-T and P-wave amplitude, frontal axis, and vector loop rotation.
• Employed EASI™ lead system transformation for redundancy features and linear support vector machine for classification.

Main Results:

• Achieved 99.9% specificity for both precordial and limb cable interchange detection.
• Overall sensitivity for limb cable interchange was 93.8%, and for precordial interchange was 56.5%.
• A low total false positive rate of 0.7% was maintained, with options for higher sensitivity.

Conclusions:

• The proposed algorithm effectively detects ECG cable interchanges with high specificity and low false positive rates.
• Sensitivity for certain precordial interchanges may be reduced, but the algorithm is configurable for different needs.
• This tool aids in improving ECG diagnostic accuracy by addressing cable misconnection errors.