The impact of imputation procedures with machine learning methods on the performance of classifiers: An application to coronary artery disease data including missing values

Background: Coronary artery disease (CAD) is an important cause of mortality and morbidity globally. Objective : The early prediction of the CAD would be valuable in identifying individuals at risk, and in focusing resources on its prevention. In this paper, we aimed to establish a diagnostic model to predict CAD by using three approaches of ANN (pattern recognition-ANN, LVQ-ANN, and competitive ANN). Methods: One promising method for early prediction of disease based on risk factors is machine learning. Among diﬀerent machine learning algorithms, the artificial neural network (ANN) algo-rithms have been applied widely in medicine and a variety of real-world classifications. ANN is a non-linear computational model, that is inspired by the human brain to analyze and process complex datasets. Results: Diﬀerent methods of ANN that are investigated in this paper indicates in both pattern recognition ANN and LVQ-ANN methods, the predictions of Angiography+ class have high accuracy. Moreover, in CNN the correlations between the individuals in cluster ”c” with the class of Angiography+ is strongly high. This accuracy indicates the significant diﬀerence among some of the input features in Angiography+ class and the other two output classes. A comparison among the chosen weights in these three methods in separating control class and Angiography+ shows that hs-CRP, FSG, and WBC are the most substantial excitatory weights in recognizing the Angiography+ individuals although, HDL-C and MCH are determined as inhibitory weights. Furthermore, the effect of decomposition of a multi-class problem to a set of binary classes and random sampling on the accuracy of the diagnostic model is investigated. Conclusion : This study confirms that pattern recognition-ANN had the most accuracy of performance among diﬀerent methods of ANN. That’s due to the back-propagation procedure of the process in which the network classify input variables based on labeled classes. The results of binarization show that decomposition of the multi-class set to binary sets could achieve higher accuracy.

Download Full-text

Impact of Coronary CT angiography in combination with CAD-RADS on the management of coronary artery disease patients

European Heart Journal - Cardiovascular Imaging ◽

10.1093/ehjci/jeaa356.237 ◽

2021 ◽

Vol 22 (Supplement_1) ◽

Author(s):

F Andre ◽

S Seitz ◽

P Fortner ◽

R Sokiranski ◽

F Gueckel ◽

...

Keyword(s):

Coronary Artery Disease ◽

Coronary Artery ◽

Risk Stratification ◽

Coronary Ct Angiography ◽

Patient Management ◽

Male Gender ◽

Coronary Ct ◽

History Of ◽

Artery Disease ◽

The Impact

Abstract Funding Acknowledgements Type of funding sources: Private company. Main funding source(s): Siemens Healthineers Introduction Coronary CT angiography (CCTA) plays an increasing role in the detection and risk stratification of patients with coronary artery disease (CAD). The Coronary Artery Disease – Reporting and Data System (CAD-RADS) allows for standardized classification of CCTA results and, thus, may improve patient management. Purpose Aim of this study was to assess the impact of CCTA in combination with CAD-RADS on patient management and to identify the impact of cardiovascular risk factors (CVRF) on CAD severity. Methods CCTA was performed on a third-generation dual-source CT scanner in patients, who were referred to a radiology centre by their attending physicians. In a total of 4801 patients, CVRF were derived from medical reports and anamnesis. Results The study population consisted of 4770 patients (62.0 (54.0-69.0) years, 2841 males) with CAD (CAD-RADS 1-5), while 31 patients showed no CAD and were excluded from further analyses. Age, male gender and the number of CVRF were associated with more severe CAD stages (all p < 0.001). 3040 patients (63.7 %) showed minimal or mild CAD requiring optimization of CVRF i.e. medical therapy but no further assessment at his time. A group of 266 patients (5.6 %) had a severe CAD defined as CAD-RADS 4B/5. In the multivariate regression analysis, age, male gender, history of smoking, diabetes mellitus and hyperlipidaemia were significant predictors for severe CAD, whereas arterial hypertension and family history of CAD did not reach significance. Of note, a subgroup of 28 patients (10.5 %) with a severe CAD (68.5 (65.5-70.0) years, 26 males, both p = n.s.) had no CVRF. Conclusions CCTA in combination with the CAD-RADS allowed for effective risk stratification of CAD patients. The majority of the patients showed non-obstructive CAD and, thus, could be treated conservatively without the need for further CAD assessment. CVRF out of arterial hypertension and family history had an impact on CAD severity reflected in higher CAD-RADs gradings. Of note, a relevant fraction of patients with CAD did not have any CVRF and, thus, may not be covered by risk stratification models. CAD-RADS n Age (years) Males (%) 1 1453 56.0 (50.0-62.0) 623 (42.9 %) 2 1587 62.0 (55.0-69.0) 918 (57.8 %) 3 1067 66.0 (59.0-71.0) 749 (70.2 %) 4A 397 66.0 (59.0-72.0) 317 (79.8 %) 4B 162 67.0 (61.0-74.0) 139 (85.8 %) 5 104 66.0 (58.5.0-77.0) 95 (91.3 %)

Download Full-text

Impact of established cardiovascular disease on 10-year death after coronary revascularization for complex coronary artery disease

Clinical Research in Cardiology ◽

10.1007/s00392-021-01922-y ◽

2021 ◽

Author(s):

Rutao Wang ◽

Scot Garg ◽

Chao Gao ◽

Hideyuki Kawashima ◽

Masafumi Ono ◽

...

Keyword(s):

Cardiovascular Disease ◽

Coronary Artery Disease ◽

Coronary Artery ◽

Coronary Revascularization ◽

Long Term Outcomes ◽

Vital Status ◽

Increased Risk ◽

Artery Disease ◽

The Impact

Abstract Aims To investigate the impact of established cardiovascular disease (CVD) on 10-year all-cause death following coronary revascularization in patients with complex coronary artery disease (CAD). Methods The SYNTAXES study assessed vital status out to 10 years of patients with complex CAD enrolled in the SYNTAX trial. The relative efficacy of PCI versus CABG in terms of 10-year all-cause death was assessed according to co-existing CVD. Results Established CVD status was recorded in 1771 (98.3%) patients, of whom 827 (46.7%) had established CVD. Compared to those without CVD, patients with CVD had a significantly higher risk of 10-year all-cause death (31.4% vs. 21.7%; adjusted HR: 1.40; 95% CI 1.08–1.80, p = 0.010). In patients with CVD, PCI had a non-significant numerically higher risk of 10-year all-cause death compared with CABG (35.9% vs. 27.2%; adjusted HR: 1.14; 95% CI 0.83–1.58, p = 0.412). The relative treatment effects of PCI versus CABG on 10-year all-cause death in patients with complex CAD were similar irrespective of the presence of CVD (p-interaction = 0.986). Only those patients with CVD in ≥ 2 territories had a higher risk of 10-year all-cause death (adjusted HR: 2.99, 95% CI 2.11–4.23, p < 0.001) compared to those without CVD. Conclusions The presence of CVD involving more than one territory was associated with a significantly increased risk of 10-year all-cause death, which was non-significantly higher in complex CAD patients treated with PCI compared with CABG. Acceptable long-term outcomes were observed, suggesting that patients with established CVD should not be precluded from undergoing invasive angiography or revascularization. Trial registration SYNTAX: ClinicalTrials.gov reference: NCT00114972. SYNTAX Extended Survival: ClinicalTrials.gov reference: NCT03417050. Graphic abstract

Download Full-text

PREDICTING MORTALITY IN PATIENTS WITH CORONARY ARTERY DISEASE REFERRED TO CARDIAC REHABILITATION: A MACHINE LEARNING APPROACH

Journal of the American College of Cardiology ◽

10.1016/s0735-1097(21)02885-0 ◽

2021 ◽

Vol 77 (18) ◽

pp. 1527

Author(s):

Christina G. De Souza E Silva ◽

Edmundo de Souza e Silva ◽

Ross Arena ◽

Jonathan Myers

Keyword(s):

Machine Learning ◽

Coronary Artery Disease ◽

Coronary Artery ◽

Cardiac Rehabilitation ◽

Learning Approach ◽

Machine Learning Approach ◽

Artery Disease

Download Full-text

Machine Learning Quantitation of Cardiovascular and Cerebrovascular Disease: A Systematic Review of Clinical Applications

Diagnostics ◽

10.3390/diagnostics11030551 ◽

2021 ◽

Vol 11 (3) ◽

pp. 551

Author(s):

Chris Boyd ◽

Greg Brown ◽

Timothy Kleinig ◽

Joseph Dawson ◽

Mark D. McDonnell ◽

...

Keyword(s):

Machine Learning ◽

Coronary Artery Disease ◽

Coronary Artery ◽

Model Comparison ◽

Disease Risk ◽

Imaging Modalities ◽

Current Status ◽

Advantages And Disadvantages ◽

Body Regions ◽

Artery Disease

Research into machine learning (ML) for clinical vascular analysis, such as those useful for stroke and coronary artery disease, varies greatly between imaging modalities and vascular regions. Limited accessibility to large diverse patient imaging datasets, as well as a lack of transparency in specific methods, are obstacles to further development. This paper reviews the current status of quantitative vascular ML, identifying advantages and disadvantages common to all imaging modalities. Literature from the past 8 years was systematically collected from MEDLINE® and Scopus database searches in January 2021. Papers satisfying all search criteria, including a minimum of 50 patients, were further analysed and extracted of relevant data, for a total of 47 publications. Current ML image segmentation, disease risk prediction, and pathology quantitation methods have shown sensitivities and specificities over 70%, compared to expert manual analysis or invasive quantitation. Despite this, inconsistencies in methodology and the reporting of results have prevented inter-model comparison, impeding the identification of approaches with the greatest potential. The clinical potential of this technology has been well demonstrated in Computed Tomography of coronary artery disease, but remains practically limited in other modalities and body regions, particularly due to a lack of routine invasive reference measurements and patient datasets.

Download Full-text

MACHINE LEARNING FOR PREDICTING ANGIOGRAPHIC PRESENCE OF ATHEROSCLEROTIC CORONARY ARTERY DISEASE IN YOUNG ADULTS

Journal of the American College of Cardiology ◽

10.1016/s0735-1097(21)01424-8 ◽

2021 ◽

Vol 77 (18) ◽

pp. 65

Author(s):

Maryam Saleem ◽

Naveena Yanamala ◽

Irfan Zeb ◽

Brijesh Patel ◽

Heenaben Patel ◽

...

Keyword(s):

Machine Learning ◽

Coronary Artery Disease ◽

Young Adults ◽

Coronary Artery ◽

Artery Disease

Download Full-text

Impact of Coronary Artery Disease and Diabetes Mellitus on the Long-Term Follow-Up in Patients after Retrograde Recanalization of the Femoropopliteal Arterial Region

Journal of Diabetes Research ◽

10.1155/2019/6036359 ◽

2019 ◽

Vol 2019 ◽

pp. 1-6 ◽

Cited By ~ 1

Author(s):

Joanna Wojtasik-Bakalarz ◽

Zoltan Ruzsa ◽

Tomasz Rakowski ◽

Andreas Nyerges ◽

Krzysztof Bartuś ◽

...

Keyword(s):

Diabetes Mellitus ◽

Coronary Artery Disease ◽

Coronary Artery ◽

Risk Of Death ◽

Significant Difference ◽

Long Term Follow Up ◽

Artery Disease ◽

The Impact

The most relevant comorbidities in patients with peripheral artery disease (PAD) are coronary artery disease (CAD) and diabetes mellitus (DM). However, data of long-term follow-up of patients with chronic total occlusion (CTO) are scarce. The aim of the study was to assess the impact of CAD and DM on long-term follow-up patients after superficial femoral artery (SFA) CTO retrograde recanalization. In this study, eighty-six patients with PAD with diagnosed CTO in the femoropopliteal region and at least one unsuccessful attempt of antegrade recanalization were enrolled in 2 clinical centers. Mean time of follow-up in all patients was 47.5 months (±40 months). Patients were divided into two groups depending on the presence of CAD (CAD group: n=45 vs. non-CAD group: n=41) and DM (DM group: n=50 vs. non-DM group: n=36). In long-term follow-up, major adverse peripheral events (MAPE) occurred in 66.6% of patients with CAD vs. 36.5% of patients without CAD and in 50% of patients with DM vs. 55% of non-DM subjects. There were no statistical differences in peripheral endpoints in both groups. However, there was a statistically significant difference in all-cause mortality: in the DM group, there were 6 deaths (12%) (P value = 0.038). To conclude, patients after retrograde recanalization, with coexisting CTO and DM, are at higher risk of death in long-term follow-up.

Download Full-text