A Dynamic Systems Approach for Detecting and Localizing of Infarct-Related Artery in Acute Myocardial Infarction...

Area of Science:

• Cardiology
• Biomedical Engineering
• Signal Processing

Background:

• Timely reperfusion therapy is crucial for improving outcomes in acute myocardial infarction (MI).
• In-hospital diagnostic testing delays can significantly impede timely treatment, impacting myocardial salvage and patient recovery.
• Accurate localization of the infarct-related artery (IRA) is essential for effective and rapid intervention.

Purpose of the Study:

• To develop and validate a novel method for the hierarchical localization of the IRA in acute MI patients using paper-based ECG.
• To assess the efficacy of a Poincare pattern ensemble-based approach incorporating multi-correlated non-stationary stochastic system dynamics for IRA localization.
• To demonstrate the potential of using short ECG recordings for rapid and accurate identification of the culprit artery.

Main Methods:

• Extraction of vectorcardiogram (VCG) diagnostic features from 2.5-second paper ECG recordings.
• Application of a Poincare pattern ensemble-based method considering multi-correlated non-stationary stochastic system dynamics.
• Validation using paper ECG records and angiograms from 106 acute MI patients and the Physionet PTB database.

Main Results:

• High accuracy in differentiating healthy controls (HC) from MI patients (97.41%).
• Accurate localization of culprit arteries (left vs. right) with 89.41 ± 9.89% accuracy.
• Distinguished between left main arteries and specific coronary arteries (RCA: 88.2 ± 11.6%, LAD vs. LCX: 93.67 ± 4.89%).

Conclusions:

• Paper ECG-based IRA localization is a feasible and accurate method for acute MI assessment.
• This technique can significantly reduce diagnostic time, enabling faster patient triage to percutaneous coronary intervention (PCI) facilities.
• The proposed method offers a valuable tool for timely management of acute coronary syndromes, potentially improving patient survival and recovery rates.