ROLE OF ARTIFICIAL INTELLIGENCE IN EARLY DETECTION OF ACUTE CORONARY SYNDROME USING ECG PATTERNS

Abstract

Background: Acute coronary syndrome (ACS) had remained a major cardiovascular emergency in which early recognition had been essential for reducing complications and mortality. Electrocardiography (ECG) had provided a rapid and widely available method for identifying ischemic changes; however, interpretation had sometimes been challenging, particularly when ECG abnormalities were subtle or atypical. Artificial intelligence (AI)-based analysis of ECG patterns had emerged as a potential approach for supporting earlier detection of ACS. Aim: This study had aimed to evaluate the role of artificial intelligence in the early detection of acute coronary syndrome using ECG patterns and to assess its diagnostic performance in comparison with conventional clinical and ECG assessment. Methods: A hospital-based observational study had been conducted at Mayo Hospital, Lahore, from October 2025 to March 2026. A total of 80 patients presenting with suspected ACS had been included. Clinical characteristics, ECG findings, and relevant laboratory investigations, including cardiac biomarkers, had been recorded. ECG patterns had been analyzed using an AI-assisted diagnostic approach, and the AI-based interpretation had been compared with the final clinical diagnosis established through standard diagnostic assessment. Measures including sensitivity, specificity, positive predictive value, and negative predictive value had been evaluated. Results: Among the 80 patients, ACS had been confirmed in 46 (57.5%) patients, while 34 (42.5%) had not been diagnosed with ACS. The AI-assisted ECG analysis had correctly identified 42 of the 46 ACS cases, corresponding to a sensitivity of 91.3%. Among patients without ACS, 29 of 34 had been correctly classified, giving a specificity of 85.3%. The positive predictive value had been 89.4%, while the negative predictive value had been 87.9%. AI-assisted ECG interpretation had demonstrated improved recognition of subtle ischemic patterns and had provided rapid risk identification compared with conventional ECG interpretation alone. Conclusion: The study had demonstrated that artificial intelligence-assisted analysis of ECG patterns had shown considerable potential for early detection of acute coronary syndrome. Its high sensitivity and relatively high specificity had indicated that AI could have served as a useful supportive tool alongside conventional clinical assessment and ECG interpretation, particularly in emergency settings. Further studies involving larger and more diverse patient populations had been required to validate these findings and establish the clinical utility of AI based ECG systems.