Atherosclerosis is one of the major reasons for cardiovascular disease including coronary heart disease, cerebral infarction and peripheral vascular disease. Atherosclerosis has no obvious symptoms in its early stages, so the key to the treatment of atherosclerosis is early intervention of risk factors. Machine learning methods have been used to predict atherosclerosis, but the presence of strong causal relationships between features can lead to extremely high levels of information redundancy, which can affect the effectiveness of prediction systems.
We aim to combine statistical analysis and machine learning methods to reduce information redundancy and further improve the accuracy of disease diagnosis.
We cleaned and collated the relevant data obtained from the retrospective study at Affiliated Hospital of Nanjing University of Chinese Medicine through data analysis. First, some features that with too many missing values are filtered out of the 34 features, leaving 25 features. 49% of the samples were categorized as the atherosclerosis risk group while the rest 51% as the control group without atherosclerosis risk under the guidance of relevant experts. We compared the prediction results of a single indicator that had been medically proven to be highly correlated with atherosclerosis with the prediction results of multiple features to fully demonstrate the effect of feature information redundancy on the prediction results. Then the features that could distinguish whether have atherosclerosis risk or not were retained by statistical tests, leaving 20 features. To reduce the information redundancy between features, after drawing inspiration from graph theory, machine learning combined with optimal correlation distances was then used to screen out 15 significant features, and the prediction models were evaluated under the 15 features. Finally, the information of the 5 screened-out non-significant features was fully utilized by ensemble learning to improve the prediction superiority for atherosclerosis.
Area Under the Receiver Operating Characteristic (ROC) Curve (AUC), which is used to measure the predictive performance of the model, was 0.84035 and Kolmogorov-Smirnov (KS) value was 0.646. After feature selection model based on optimal correlation distance, the AUC value was 0.88268 and the KS value was 0.688, both of which were improved by about 0.04. Finally, after ensemble learning, the AUC value of the model was further improved by 0.01369 to 0.89637.
The optimal distance feature screening model proposed in this paper improves the performance of atherosclerosis prediction models in terms of both prediction accuracy and AUC metrics. Code and models are available at https://github.com/Cesartwothousands/Prediction-of-Atherosclerosis.
Citation: Zihan Chen, Minhui Yang, Yuhang Wen, Songyan Jiang, Wenjun Liu, Hui Huang. Prediction of atherosclerosis using machine learning based on operations research[J]. Mathematical Biosciences and Engineering, 2022, 19(5): 4892-4910. doi: 10.3934/mbe.2022229
Atherosclerosis is one of the major reasons for cardiovascular disease including coronary heart disease, cerebral infarction and peripheral vascular disease. Atherosclerosis has no obvious symptoms in its early stages, so the key to the treatment of atherosclerosis is early intervention of risk factors. Machine learning methods have been used to predict atherosclerosis, but the presence of strong causal relationships between features can lead to extremely high levels of information redundancy, which can affect the effectiveness of prediction systems.
We aim to combine statistical analysis and machine learning methods to reduce information redundancy and further improve the accuracy of disease diagnosis.
We cleaned and collated the relevant data obtained from the retrospective study at Affiliated Hospital of Nanjing University of Chinese Medicine through data analysis. First, some features that with too many missing values are filtered out of the 34 features, leaving 25 features. 49% of the samples were categorized as the atherosclerosis risk group while the rest 51% as the control group without atherosclerosis risk under the guidance of relevant experts. We compared the prediction results of a single indicator that had been medically proven to be highly correlated with atherosclerosis with the prediction results of multiple features to fully demonstrate the effect of feature information redundancy on the prediction results. Then the features that could distinguish whether have atherosclerosis risk or not were retained by statistical tests, leaving 20 features. To reduce the information redundancy between features, after drawing inspiration from graph theory, machine learning combined with optimal correlation distances was then used to screen out 15 significant features, and the prediction models were evaluated under the 15 features. Finally, the information of the 5 screened-out non-significant features was fully utilized by ensemble learning to improve the prediction superiority for atherosclerosis.
Area Under the Receiver Operating Characteristic (ROC) Curve (AUC), which is used to measure the predictive performance of the model, was 0.84035 and Kolmogorov-Smirnov (KS) value was 0.646. After feature selection model based on optimal correlation distance, the AUC value was 0.88268 and the KS value was 0.688, both of which were improved by about 0.04. Finally, after ensemble learning, the AUC value of the model was further improved by 0.01369 to 0.89637.
The optimal distance feature screening model proposed in this paper improves the performance of atherosclerosis prediction models in terms of both prediction accuracy and AUC metrics. Code and models are available at https://github.com/Cesartwothousands/Prediction-of-Atherosclerosis.
| [1] |
C. Sinning, A. Kieback, P. S. Wild, R. B. Schnabel, F. Ojeda, S. Appelbaum, et al., Association of multiple biomarkers and classical risk factors with early carotid atherosclerosis: results from the Gutenberg Health Study, Clin. Res. Cardiol., 103 (2014), 477-485. https://doi.org/10.1007/s00392-014-0674-6 doi: 10.1007/s00392-014-0674-6
|
| [2] | J. F. Polak, M. J. Pencina, D. H. O'Leary, R. B. D'Agostino, Common carotid artery intima-media thickness progression as a predictor of stroke in multi-ethnic study of atherosclerosis, Stroke, 42 (2011), 3017-3021. https://doi.org/10.1161/STROKEAHA.111.625186 |
| [3] |
M. W. Lorenz, C. Schaefer, H. Steinmetz, M. Sitzer, Is carotid intima media thickness useful for individual prediction of cardiovascular risk? Ten-year results from the Carotid Atherosclerosis Progression Study (CAPS), Eur. Heart J., 31 (2010), 2041-2048. https://doi.org/10.1093/eurheartj/ehq189 doi: 10.1093/eurheartj/ehq189
|
| [4] | M. Soni, M. Ambrosino, D. S. Jacoby, The use of subclinical atherosclerosis imaging to guide preventive cardiology management, Curr. Cardiol. Rep., 23 (2021), 61. https://doi.org/10.1007/s11886-021-01490-7 |
| [5] | A. Hazra, S. K. Mandal, A. Gupta, A. Mukherjee, A. Mukherjee, Heart disease diagnosis and prediction using machine learning and data mining techniques: a review, Adv. Comput. Sci. Technol., 10 (2017), 2137-2159. |
| [6] |
M. Shouman, T. Turner, R. Stocker, Integrating Naive Bayes and K-means clustering with different initial centroid selection methods in the diagnosis of heart disease patients, Comput. Sci. Conf. Proc., 5 (2012), 125-137. https://doi.org/10.5121/csit.2012.2511 doi: 10.5121/csit.2012.2511
|
| [7] | O. Terrada, B. Cherradi, A. Raihani, O. Bouattane, Classification and prediction of atherosclerosis diseases using machine learning algorithms, in International Conference on Optimization and Applications (ICOA), 5 (2019), 1-5. https://doi.org/10.1109/ICOA.2019.8727688 |
| [8] | D. Han, K. K. Kolli, S. J. Al'Aref, L. Baskaran, A. R. van Rosendael, H. Gransar, et al., Machine learning framework to identify individuals at risk of rapid progression of coronary atherosclerosis: from the PARADIGM registry, J. Am. Heart Assoc., 9 (2020), e013958. https://doi.org/10.1161/JAHA.119.013958 |
| [9] | O. Couturier, H. Delalin, H. Fu, G. Edouard, A three-step approach for stulong database analysis: characterization of patients groups, in Proceeding of the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery, 2004. |
| [10] |
M. Abdar, W. Ksiazek, U. R. Acharya, R. S. Tan, V. Makarenkov, P. Plawiak, A new machine learning technique for an accurate diagnosis of coronary artery disease, Comput. Methods Programs Biomed., 179 (2019), 104992. https://doi.org/10.1016/j.cmpb.2019.104992 doi: 10.1016/j.cmpb.2019.104992
|
| [11] |
V. S. H. Rao, M. N. Kumar, Novel approaches for predicting risk factors of atherosclerosis, IEEE J. Biomed. Health, 17 (2012), 183-189. https://doi.org/10.1109/TITB.2012.2227271 doi: 10.1109/TITB.2012.2227271
|
| [12] | J. Xie, R. Wu, H. Wang, Y. Kong, H. Li, W. Zhang, A novel weight learning approach based on density for accurate prediction of atherosclerosis, in Intelligent Computing Theories and Application (eds. D. S. Huang, K. H. Jo., Z. K. Huang), Springer, (2019), 190-200. https://doi.org/10.1007/978-3-030-26969-2_18 |
| [13] |
W. He, Y. Xie, H. Lu, M. Wang, H. Chen, Predicting coronary atherosclerotic heart disease: an extreme learning machine with improved salp swarm algorithm, Symmetry, 12 (2020), 1651. https://doi.org/10.3390/sym12101651 doi: 10.3390/sym12101651
|
| [14] |
A. Ward, A. Sarraju, S. Chung, J. Li, R. Harrington, P. Heidenreich, Machine learning and atherosclerotic cardiovascular disease risk prediction in a multi-ethnic population, NPJ Digit. Med., 125 (2020), 1-7. https://doi.org/10.1038/s41746-020-00331-1 doi: 10.1038/s41746-020-00331-1
|
| [15] | S. Nikan, F. Gwadry-Sridhar, M. Bauer, Machine learning application to predict the risk of coronary artery atherosclerosis, in 2016 International Conference on Computational Science and Computational Intelligence (CSCI), (2016), 34-39. https://doi.org/10.1109/CSCI.2016.0014 |
| [16] | J. Xie, H. Wang, J. Zhang, C. Meng, Y Kong, S. Mao, et al., A novel hybrid subset-learning method for predicting risk factors of atherosclerosis, in 2017 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), (2017), 2124-2131. https://doi.org/10.1109/BIBM.2017.8217987 |
| [17] |
M. Priya, P. Ranjith Kumar, A novel intelligent approach for predicting atherosclerotic individuals from big data for healthcare, Int. J. Prod. Res., 53 (2015), 7517-7532. https://doi.org/10.1080/00207543.2015.1087655 doi: 10.1080/00207543.2015.1087655
|
| [18] | A. I. Sakellarios, V. C. Pezoulas, C. Bourantas, K. K. Naka, L. K. Michalis, P. W. Serruys, et al., Prediction of atherosclerotic disease progression combining computational modelling with machine learning, in 2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), (2020), 2760-2763. https://doi.org/10.1109/EMBC44109.2020.9176435 |
| [19] | B. Kumar, H. Mathur, Comprehensive analysis of atherosclerosis disease prediction using machine learning, Ann. Rom. Soc. Cell Biol., 4 (2021), 17962-17975. |
| [20] |
M. Lin, H. Cui, W. Chen, A. van Engelen, M. de Bruijne, M. R. Azarpazhooh, et al., Longitudinal assessment of carotid plaque texture in three-dimensional ultrasound images based on semi-supervised graph-based dimensionality reduction and feature selection, Comput. Biol. Med., 116 (2020), 103586. https://doi.org/10.1016/j.compbiomed.2019.103586 doi: 10.1016/j.compbiomed.2019.103586
|
| [21] |
Q. A. Hathaway, N. Yanamala, M. J. Budoff, P. P. Sengupta, I. Zeb, Deep neural survival networks for cardiovascular risk prediction: The Multi-Ethnic Study of Atherosclerosis (MESA), Comput. Biol. Med., 139 (2021), 104983. https://doi.org/10.1016/j.compbiomed.2021.104983 doi: 10.1016/j.compbiomed.2021.104983
|
| [22] |
A. D. Jamthikar, D. Gupta, L. Saba, N. N. Khanna, K. Viskovic, S. Mavrogeni, et al., Artificial intelligence framework for predictive cardiovascular and stroke risk assessment models: A narrative review of integrated approaches using carotid ultrasound, Comput. Biol. Med., 126 (2020), 104043. https://doi.org/10.1016/j.compbiomed.2020.104043 doi: 10.1016/j.compbiomed.2020.104043
|
| [23] |
S. S. Skandha, S. K. Gupta, L. Saba, V. K. Koppula, A. M. Johri, N. N. Khanna, et al., 3-D optimized classification and characterization artificial intelligence paradigm for cardiovascular/stroke risk stratification using carotid ultrasound-based delineated plaque: AtheromaticTM 2.0, Comput. Biol. Med., 125 (2020), 103958. https://doi.org/10.1016/j.compbiomed.2020.103958 doi: 10.1016/j.compbiomed.2020.103958
|
| [24] | R. H. Lopes, I. D. Reid, P. R. Hobson, The two-dimensional Kolmogorov-Smirnov test, Prod. Sci., (2007), 1-12. |
| [25] | G. Biau, E.Scornet, A random forest guided tour, Test, 25 (2016), 197-227. https://doi.org/10.1007/s11749-016-0481-7 |
| [26] | M. Noto, H. Sato, A method for the shortest path search by extended Dijkstra algorithm, in Smc 2000 conference proceedings. 2000 ieee international conference on systems, man and cybernetics. 'cybernetics evolving to systems, humans, organizations, and their complex interactions', 3 (2000), 2316-2320. https://doi.org/10.1109/ICSMC.2000.886462 |
| [27] | N. Friedman, D. Geiger, M. Goldszmidt, Bayesian network classifiers, Mach. Learn., 29 (1997), 131-163. https://doi.org/10.1023/A:1007465528199 |
| [28] |
F. Xu, J. Zhang, X. Zhou, H. Hao, Lipoxin A4 and its analog attenuate high fat diet-induced atherosclerosis via Keap1/Nrf2 pathway, Exp. Cell Res., 412 (2022), 113025. https://doi.org/10.1016/j.yexcr.2022.113025 doi: 10.1016/j.yexcr.2022.113025
|
| [29] | F. Polak, J. Y. C. Backlund, M. Budoff, P. Raskin, I. Bebu, J. M. Lachin, et al., Coronary artery disease events and carotid intima-media thickness in Type 1 diabetes in the DCCT/EDIC cohort, J. Am. Heart Assoc., 24 (2021), e022922. https://doi.org/10.1161/JAHA.121.022922 |
Figures(8) / Tables(3)
Zihan Chen, Minhui Yang, Yuhang Wen, Songyan Jiang, Wenjun Liu, Hui Huang. Prediction of atherosclerosis using machine learning based on operations research[J]. Mathematical Biosciences and Engineering, 2022, 19(5): 4892-4910. doi: 10.3934/mbe.2022229
DownLoad: