TP Abdualimov¹* and AG Obrezan^1,2

¹Department of General Practice and Therapy, My Medical Center, Saint-Petersburg, Russian Federation
²Department of Hospital Therapy, St. Petersburg State University, Saint-Petersburg, Russian Federation

*Corresponding Author: TP Abdualimov, Department of General Practice and Therapy, My Medical Center, Saint-Petersburg, Russian Federation.

Received: January 27, 2023; Published: February 14, 2023

Abstract

The purpose of the study is to study the possibility of using neural network analysis to predict the fact and degree of coronary lesion. The task of the study was to compare the accuracy of the trained neural network model on input structured data and ECG records with the conclusion of a cardiologist.

Material and methods: 120 patients who underwent elective or emergency coronary angiography. To predict the damage to the coronary bed, the method of neural network analysis was used. Machine learning was performed with the inclusion of clinical, laboratory, instrumental (ECG-image) signs (23 indicators in total). To solve classification problems, a neural network was used that takes structured data and an image as input and outputs a multifactorial characteristic of the main coronary arteries.

The training/test ratio in the examples was 100/20. The supervised learning method was used on the available data, in which the outcomes were known (coronary angiography data), and the neural network parameters were adjusted so as to minimize the error using the backpropagation method. For this experiment, based on the test sample, 20 tasks were created. An ECG image was attached to each task.

5 cardiologists, daily supervising patients with acute coronary syndrome, separately assessed the pathology of the coronary bed for each main coronary artery and predicted the need for revascularization.

Results: The results of the neural network on a test sample of 20 patients: AUC score - 0.74, accuracy (accuracy) - 80%, precision accuracy (precision) - 63%, recall (recall) - 55%, f1 score (harmonic mean between accuracy and completeness) - 59%. Average response rates of cardiologists: accuracy - 76%, precision - 48%, recall - 55%, AUC score - 0.68, f1 score - 49%. The best values of cardiologists: accuracy - 76%, precision - 48%, recall - 67%, AUC score - 0.72, f1 score - 56%.

Conclusion: Neural network analysis of the prepared clinical, laboratory and instrumental data allows you to adjust the network parameters for subsequent accurate prediction of coronary artery disease. The results obtained in the form of an AUC score allow us to speak about the applicability of the method in the diagnosis of coronary pathology. On the test sample, the neural network works more efficiently than cardiologists on average. Only one in five specialists was able to come close to the accuracy of the trained neural network model.

Keywords: Coronary Arteries; Neural Networks; Artificial Intelligence; Coronary Heart Disease; Deep Learning; ECG; Non-Invasive Predictive AI Coronary Angiography

References

1. Kuo FC., et al. “The relative utilities of genome-wide, gene panel, and individual gene sequencing in clinical practice”. Blood 130 (2017): 433-439.
2. Muse ED., et al. “Towards a smart medical home”. Lancet 389 (2017): 358.
3. Steinhubl SR., et al. “The emerging field of mobile health”. Science Translational Medicine 7 (2015): 283rv3.
4. Shameer K., et al. “Translational bioinformatics in the era of real-time biomedical, health care and wellness data streams”. Brief Bioinformation 18 (2017): 105-124.
5. Konstam MA., et al. “The academic medical system: reinvention to survive the revolution in health care”. Journal of the American College of Cardiology 69 (2017): 1305-1312.
6. Steinhubl SR and Topol EJ. “Moving from digitalization to digitization in cardiovascular care: why is it important, and what could it mean for patients and providers?” Journal of the American College of Cardiology 66 (2015): 1489-1496.
7. Boeldt DL., et al. “How consumers and physicians view new medical technology: comparative survey”. Journal of Medical Internet Research 17 (2015): e215.
8. Wasserlauf J., et al. “Smartwatch performance for the detection and quantification of atrial fibrillation”. Circulation: Arrhythmia and Electrophysiology 12 (2019): e006834.

Citation

Citation: TP Abdualimov and AG Obrezan. “Possibilities of Artificial Intelligence in Predicting Coronary Artery Disease”.Acta Scientific Medical Sciences 7.3 (2023): 104-108.

Copyright

Copyright: © 2022 TP Abdualimov and AG Obrezan. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.