Inteligência artificial na estratificação de risco da síndrome coronariana aguda:  uma revisão comparativa com escores clínicos tradicionais

Suzan Kelly  Macedo; Bruna  de Souza Ferreira; Karoline  José de Deus Souza Gomes; Guilherme  Tomas Luciano; Fabiano Inácio  de Souza

Authors

Suzan Kelly Macedo Graduando em Medicina na Universidade Federal de Goiás (UFG), Goiânia/GO, Brasil. Autor https://orcid.org/0009-0004-8978-7387
Bruna de Souza Ferreira Graduando em Medicina na Universidade Federal de Goiás (UFG), Goiânia/GO, Brasil. Autor https://orcid.org/0009-0004-9240-9838
Karoline José de Deus Souza Gomes Graduanda em Medicina no Pontifício Universidade Católica de Goiás, Goiânia/GO, Brasil. Autor https://orcid.org/0009-0009-0967-1656
Guilherme Tomas Luciano Graduando em Medicina na Universidade Federal de Goiás (UFG), Goiânia/GO, Brasil. Autor https://orcid.org/0009-0002-9281-0897
Fabiano Inácio de Souza Pós Doutorando em Medicina na Universidade Federal de Goiás (UFG), Goiânia/GO, Brasil Autor https://orcid.org/0000-0001-5862-4674

Keywords:

Machine learning, Cardiology, Artificial intelligence, Risk assessment, Acute coronary syndrome

Abstract

Acute Coronary Syndrome (ACS) demands early and effective risk stratification to guide rapid clinical decisions in emergency settings. Given the limitations of traditional scores in capturing complex non-linear biological interactions, this study aimed to review the discriminatory and predictive performance of Artificial Intelligence (AI) models, such as Machine Learning and Deep Learning, compared to conventional tools (GRACE, TIMI, and HEART) in predicting Major Adverse Cardiovascular Events (MACE). For this purpose, an integrative literature review was conducted based on the PRISMA 2020 protocol, selecting nine original articles published in PubMed between 2021 and 2026. The results consistently demonstrate that AI algorithms outperform the accuracy of current clinical scores, achieving Areas Under the ROC Curve (AUROC) of up to 0.94. This prognostic refinement is rooted in the autonomous extraction of subtle morphological variables from raw electrocardiograms — drastically reducing false positives in the cath lab — and the synergistic integration of molecular and inflammatory biomarkers. Explainable tools such as SHAP validate mathematical findings by aligning them with medical guidelines. In conclusion, AI represents a disruptive advancement in ACS risk stratification. However, definitive bedside translation still depends on overcoming methodological challenges, requiring external validations with real-time live data streams to establish a safe and personalized clinical practice.

References

Petrescu L, Adverse Events in Acute Coronary Syndrome Authors, et al. Artificial Intelligence and Its Role in the Diagnosis and Prediction of Adverse Events in Acute Coronary Syndrome: A Narrative Review of the Literature. Life (Basel) [Internet]. 2025 [Citado em 1 Jun 2026]; 15(1): 1-15. Disponível em: https://www.mdpi.com/journal/life. doi: 10.3390/life15010001.

Meyer A, Stratification Authors, et al. Artificial intelligence in risk stratification of acute coronary syndrome: Future vision or already reality?. Herz [Internet]. 2026 [Citado em 1 Jun 2026]; 51(2): 112-120. Disponível em: https://link.springer.com/journal/59. doi: 10.1007/s00059-026-05432-1.

Elsayed M, Risk Calculator Authors, et al. Development and Technical Validation of an Integrated Risk Calculator for Acute Coronary Syndrome Using ChatGPT-Assisted Coding. Cureus [Internet]. 2025 [Citado em 1 Jun 2026]; 17(3): e40462. Disponível em: https://www.cureus.com. doi: 10.7759/cureus.40462.

Commandeur F, Pooled Analysis Authors, et al. Machine learning versus traditional risk stratification methods in acute coronary syndrome: a pooled randomized clinical trial analysis. J Thromb Haemost. 2020; 18(4): 852-863. doi: 10.1111/jth.14725.

Zhang X, ECG Suspected ACS Authors, et al. Electrocardiogram-based machine learning for risk stratification of patients with suspected acute coronary syndrome. Eur Heart J Digit Health [Internet]. 2025 [Citado em 1 Jun 2026]; 6(1): ztae091. Disponível em: https://academic.oup.com/ehjdh. doi: 10.1093/ehjdh/ztae091.

Song L, Li Y, Nie S, Feng Z, Liu Y, Ding F, Gong L, Liu L, Yang G. Using machine learning to predict adverse events in acute coronary syndrome: A retrospective study. Clin Cardiol [Internet]. Dez 2023 [Citado em 1 Jun 2026]; 46(12): 1594-1602. Disponível em: https://onlinelibrary.wiley.com/journal/10974598. doi: 10.1002/clc.24127.

Kisova T, Demolder A, Rafajdus A, Iring A, Palus T, Belmonte M, Barbato E, Baron SJ, Hatala R, Smith SW, Meyers HP, Sharkey SW, Bartunek J, Henry TD. AI-Enabled ECG Analysis Improves Diagnostic Accuracy and Reduces False STEMI Activations: A Multicenter U.S. Registry. JACC Cardiovasc Interv. 2026 Jan 26; 19(2): 145-156. doi: 10.1016/j.jcin.2025.10.018.

Bouzid Z, Sejdic E, Martin-Gill C, Faramand Z, Frisch S, Alrawashdeh M, Helman S, Gokhale TA, Riek NT, Kraevsky-Phillips K, Gregg RE, Sereika SM, Clermont G, Akcakaya M, Zègre- Hemsey JK, Saba S, Callaway CW, Al-Zaiti SS. Electrocardiogram-based machine learning for risk stratification of patients with suspected acute coronary syndrome. Eur Heart J. 7 Mar 2025; 46(10): 943-954. doi: 10.1093/eurheartj/ehae880.

Jalepalli SK, Gupta P, Dekker ALAJ, Bermejo I, Kar S. Development and validation of multicentre study on novel Artificial Intelligence-based Cardiovascular Risk Score (AICVD). Fam Med Community Health [Internet]. 18 Jan 2024 [Citado em 1 Jun 2026].

Haimovich JS, Kolossváry M, Alam R, Padrós-Valls R, Lu MT, Aguirre AD. Risk stratification of chest pain in the emergency department using artificial intelligence applied to electrocardiograms. Open Heart [Internet]. 1 Set 2025 [Citado em 1 Jun 2026]; 12(2): e003343. Disponível em: https://openheart.bmj.com. doi: 10.1136/openhrt-2025-003343.

Jung JH, Lee K, Chang K, Ahn Y, Her SH, Lee S. Comparison of the Prognostic Performance of Various Machine Learning Models in Patients with Acute Myocardial Infarction: Results from the COREA-AMI Registry. Medicina (Kaunas) [Internet]. 2 Out 2025 [Citado em 1 Jun 2026]; 61(10): 1783. Disponível em: https://www.mdpi.com/1648-9144/61/10/1783.

Perrone A, Khoshgoftaar TM. A review of machine learning applications in heart health. Biomed Eng Online [Internet]. 11 Ago 2025 [Citado em 1 Jun 2026]; 24(1): 99. Disponível em: https://biomedical-engineering-online.biomedcentral.com. doi: 10.1186/s12938-025-01430-4.

Stătescu C, Anghel L, Tudurachi BS, Leonte A, Benchea LC, Sascău RA. From Classic to Modern Prognostic Biomarkers in Patients with Acute Myocardial Infarction. Int J Mol Sci [Internet]. 15 Ago 2022 [Citado em 1 Jun 2026]; 23(16): 9168. Disponível em: https://www.mdpi.com/journal/ijms. doi: 10.3390/ijms23169168.

Roseiro M, Henriques J, Paredes S, Rocha T, Sousa J. An interpretable machine learning approach to estimate the influence of inflammation biomarkers on cardiovascular risk assessment. Comput Methods Programs Biomed. Mar 2023; 230: 107347. doi: 10.1016/j.cmpb.2023.107347.

Sousa S, Paredes S, Rocha T, Henriques J, Sousa J, Gonçalves L. Machine learning models’ assessment: trust and performance. Med Biol Eng Comput [Internet]. Nov 2024 [Citado em 1 Jun 2026]; 62(11): 3397-3410. Disponível em: https://link.springer.com/journal/11517. doi: 10.1007/s11517-024-03145-5.

PAGE, Matthew J. et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ, [s. l.], v. 372, n. 71, p. 1-9, 29 mar. 2021. Disponível em: https:// doi.org/10.1136/bmj.n71. Acesso em: 12 jun. 2026.