2024, issue 4, p. 71-80
Received 24.09.2024; Revised 27.10.2024; Accepted 03.12.2024
Published 18.12.2024; First Online 23.12.2024
https://doi.org/10.34229/2707-451X.24.4.7
Previous | FULL TEXT (in Ukrainian) | Next
Prediction and Assessment of Myocardial Infarction Risk on the Base of Medical Report Text Collection
Margaryta Prazdnikova
V.M. Glushkov Institute of Cybernetics of the NAS of Ukraine, Kyiv
Correspondence: This email address is being protected from spambots. You need JavaScript enabled to view it.
Introduction. Myocardial infarction remains one of the leading causes of death worldwide, resulting from sudden disruption of blood supply to the heart muscle. Key risk factors include smoking, age, gender, high cholesterol levels, diabetes, and others. Despite advancements in diagnostics and treatment, early detection of heart attack risk is crucial for reducing mortality and improving patient quality of life. This paper explores an approach to predicting heart attack risk based on analysis of text data of medical reports using machine learning.
The purpose of the article is to demonstrate how the application of machine learning, particularly the Naive Bayes classifier, can enhance the prediction of myocardial infarction risk through the analysis of extensive medical data. By leveraging a depersonalized database from SSO CITHC SAA, containing medical records collected during a decade of operating, this study seeks to reveal how the identification of critical patterns and factors can improve prediction accuracy. Additionally, the article explores how integrating these predictive models into clinical decision support systems can refine medical diagnostics and decision-making processes.
Results. The proposed prediction model demonstrated high efficiency in identifying patients at increased risk of heart attack. By analyzing the frequency of specific words in medical records, the algorithm successfully predicted a high risk of heart attack for 80 % of patients with an expected event. This underscores the significant potential of leveraging textual data and machine learning methods in medical diagnostics. Moreover, the reduction in false predictions highlights the model's reliability and suitability for practical application.
Conclusions. Employing machine learning for heart attack risk prediction based on medical data analysis represents a promising direction in modern medicine. The developed model showcases the possibility of enhancing diagnostic and predictive accuracy, which can substantially influence treatment strategy decisions and improve patient outcomes. Integrating such tools into clinical practice will facilitate more informed decisions by physicians and reduce patient risks.
Keywords: myocardial infarction, risk prediction, machine learning, database, Naive Bayes classifier, medical analytics.
Cite as: Prazdnikova M. Prediction and Assessment of Myocardial Infarction Risk on the Base of Medical Report Text Collection. Cybernetics and Computer Technologies. 2024. 4. P. 71–80. (in Ukrainian) https://doi.org/10.34229/2707-451X.24.4.7
References
1. Cabinet of Ministers of Ukraine. How to Detect a Heart Attack and What to Do in the Event of a Cardiac Arrest: Expert Comments. July 24, 2024. (in Ukrainian) https://kmu.gov.ua/news/yak-viyaviti-infarkt-i-shcho-robiti-pri-sercevomu-napadi-komentuye-ekspert (accessed: 24.07.2024)
2. Bemando C., Miranda E., Aryuni M. Machine-Learning-Based Prediction Models of Coronary Heart Disease Using Naïve Bayes and Random Forest Algorithms. International Conference on Software Engineering & Computer Systems and 4th International Conference on Computational Science and Information Management (ICSECS-ICOCSIM), Pekan, Malaysia, 2021. P. 232–237. https://doi.org/10.1109/ICSECS52883.2021.00049
3. Nandal N., Goel L., Tanwar R. Machine learning-based heart attack prediction: A symptomatic heart attack prediction method and exploratory analysis. F1000Research. 2022. 11:1126. https://doi.org/10.12688/f1000research.123776.1
4. Park J., Kim J., Kang S.-H., Lee J., Hong Y., Chang H.-J., et al. Artificial intelligence–enhanced electrocardiography analysis as a promising tool for predicting obstructive coronary artery disease in patients with stable angina. European Heart Journal - Digital Health. Vol. 5, Iss. 4. P. 444–453. https://doi.org/10.1093/ehjdh/ztae038
5. Medvediev M.G. Levenshtein Distance and Related Problems. Scientific Notes of NaUKMA. Computer Science. 2000. Vol. 18. P. 33–37. (in Ukrainian). https://ekmair.ukma.edu.ua/server/api/core/bitstreams/c1190351-8225-469b-9d01-3eabafb6fee4/content
6. Jain Y. Principal Component Analysis (PCA) [NLP, Python]. 2024. https://medium.com/@yashj302/principal-component-analysis-pca-nlp-python-ce9caa58bd7a (accessed: 09.09.2024)
7. Suman S. Text Data Pre-Processing Using Word2Vector and t-SNE. 2024. https://medium.com/swlh/text-data-pre-processing-using-word2vector-and-t-sne-2321fbce5b9 (accessed: 09.09.2024)
8. Becht E., Dutertre C.-A., Kwok I.W.H., Ng L.G., Ginhoux F., Newell E.W. Evaluation of UMAP as an alternative to t-SNE for single-cell data. The preprint server of biology. April 10, 2018. https://doi.org/10.1101/298430
9. Nigam K., McCallum A., Thrun S., Mitchell T. Learning to classify text from labeled and unlabeled documents using EM. Machine Learning. 2000. 39. P. 103–134. https://doi.org/10.1023/A:1007692713085
10. Prazdnikova M.O., Kravchenko A.M., Tulchynsky V.G., Chaikovsky I.A. Using the Database of the State Scientific Center for Preventive and Personalized Cardiology, State Administration of Affairs, for Myocardial Infarction Risk Prediction: From Concept to Implementation. International Scientific and Practical Conference "Organizational and Clinical Aspects of a Patient-Oriented Approach to Treatment and Rehabilitation in Modern Conditions", 29-30 May 2024. P. 152. (in Ukrainian) https://ua-medical.com/downloads/conferences/abstracts-2024.pdf
ISSN 2707-451X (Online)
ISSN 2707-4501 (Print)
Previous | FULL TEXT (in Ukrainian) | Next