Myocardial Ischemia Diagnosis Using a Reduced Lead System

Area of Science:

• Cardiology
• Biomedical Engineering
• Statistical Modeling

Background:

• Acute myocardial infarction (AMI) diagnosis relies heavily on electrocardiogram (ECG) interpretation.
• Standard 12-lead ECG analysis can be complex, prompting research into simplified diagnostic approaches.
• Identifying novel ECG features for improved AMI detection is crucial for timely intervention.

Purpose of the Study:

• To develop and validate a novel statistical model for diagnosing AMI.
• To evaluate the diagnostic performance of a reduced three-lead ECG system.
• To assess the utility of pseudo-vectorcardiography (pseudo-VCG) parameters in AMI diagnosis.

Main Methods:

• A statistical model was developed using features from a reduced three-lead ECG system.
• Features included standard clinical parameters (ST elevation, T-wave maximum) and novel pseudo-VCG parameters.
• The model was validated on 104 patients from the Physionet STAFF III database, including recordings during percutaneous coronary angioplasty (PTCA).

Main Results:

• Pseudo-VCG features alone showed slightly superior diagnostic performance (AUC 0.87) compared to standard features (AUC 0.85).
• Combining pseudo-VCG features with ST elevation and T-wave maximum significantly improved diagnostic performance (AUC 0.95, sensitivity 89.6%, specificity 82.7%).

Conclusions:

• The proposed reduced lead system and feature set demonstrate significant potential for accurate AMI diagnosis.
• This model could facilitate earlier and faster AMI diagnosis in long-term, ambulatory, and home monitoring scenarios.
• The integration of pseudo-VCG parameters offers a promising advancement in ECG-based cardiac diagnostics.