Comparing performances of logistic regression, classification and regression tree, and neural networks for predicting coronary artery disease

2008 ◽  
Vol 34 (1) ◽  
pp. 366-374 ◽  
Author(s):  
Imran Kurt ◽  
Mevlut Ture ◽  
A. Turhan Kurum
Keyword(s):  
Neural Networks ◽  
Coronary Artery Disease ◽  
Logistic Regression ◽  
Coronary Artery ◽  
Regression Tree ◽  
Classification And Regression Tree ◽  
Classification And Regression ◽  
Artery Disease

PHACTR1 gene polymorphism with the risk of coronary artery disease in Chinese Han population

2019 ◽  
Vol 95 (1120) ◽  
pp. 67-71 ◽  
Author(s):  
Lishan Chen ◽  
Hang Qian ◽  
Zhihuan Luo ◽  
Dongfeng Li ◽  
Hao Xu ◽  
...  
Keyword(s):  
Coronary Artery Disease ◽  
Logistic Regression ◽  
Coronary Artery ◽  
Genetic Model ◽  
Smoking Habit ◽  
Chinese Han Population ◽  
Chinese Han ◽  
Han Population ◽  
Increased Risk ◽  
Artery Disease

BackgroundCoronary artery disease (CAD) is the most frequent multifactorial disease worldwide and is characterised by endothelial injury, lipid deposition and coronary artery calcification. The purpose of this study was to determine the allelic and genotypic frequencies of two loci (rs2026458 and rs9349379) of phosphatase and actin regulator 1 (PHACTR1) to the risk of developing CAD in the Chinese Han population.MethodsA case–control study was conducted including 332 patients with CAD and 119 controls. Genotype analysis was performed by PCR and Sanger sequencing. Genetic model analysis was performed to evaluate the association between single nucleotide polymorphisms and CAD susceptibility using Pearson’s χ2 test and logistic regression analysis.ResultsThe GG genotype of rs9349379 represented 50% and 29% of patients with CAD and controls, respectively (p<0.001). The CC genotype of rs2026458 was more prevalent in the controls than in patients with CAD compared with TT genotype (OR=0.548, 95% CI 0.351 to 0.856, p=0.008). Logistic regression analyses revealed that PHACTR1 rs9349379 GG genotype was significantly associated with increased risk of CAD in the recessive model (OR=2.359, 95% CI 1.442 to 3.862, p=0.001), even after adjusting for age gender, hypertension, type 2 diabetes, hyperlipidaemia and smoking habit. Heterogeneity test proved that rs9349379’s risk effects on CAD were more significant among women.ConclusionsOur study indicate that the PHACTR1 rs9349379 polymorphism is associated with the increased risk for CAD in the female Chinese Han population.

Circulating t-PA antigen predicts major adverse coronary events in patients with stable coronary artery disease – a 13-year follow-up

2003 ◽  
Vol 90 (08) ◽  
pp. 344-350 ◽  
Author(s):  
Alexander Niessner ◽  
Senta Graf ◽  
Mariam Nikfardjam ◽  
Walter Speidl ◽  
Renate Huber-Beckmann ◽  
...  
Keyword(s):  
Insulin Resistance ◽  
Coronary Artery Disease ◽  
Logistic Regression ◽  
Coronary Artery ◽  
Predictive Value ◽  
Odds Ratio ◽  
Coronary Events ◽  
Artery Disease ◽  
Prognostic Power

SummaryThrombus formation after rupture of an atherosclerotic plaque plays a crucial role in coronary artery disease (CAD). A decreased endogenous fibrinolytic system and prothrombotic factors are supposed to influence coronary thrombosis. It was our aim to investigate the predictive value of tissue plasmino-gen activator (t-PA) antigen, von Willebrand Factor, Lipoprotein (a) and anti-cardiolipin antibodies for major adverse coronary events in patients with stable CAD in a prospective cohort study of more than 10 years.We observed 141 patients with angiographically proven CAD for a median follow-up period of 13 years. t-PA antigen was the only marker predicting coronary events (logistic regression, p = 0.044) with a poor prognosis for patients in the 5th quintile with an odds ratio of 7.3 (compared to the 1st quintile). The odds ratio even increased to 10.0 for coronary events associated with the “natural course” of CAD excluding events due to restenosis. t-PA antigen had a slightly higher prognostic power (ROC curve; AUC = 0.69) than fasting glucose (AUC = 0.68) and cholesterol (AUC = 0.67). Triglycerides influenced plasma levels of t-PA antigen (regression, p < 0.001). The predictive value of t-PA antigen remained significant after adjustment for inflammation (logistic regression, p = 0.013) and extent of CAD (p = 0.045) but disappeared adjusting for insulin resistance (p = 0.12).In conclusion t-PA antigen predicted coronary events during a very long-term follow-up with a comparable prognostic power to established cardiovascular risk factors. Markers of insulin resistance influenced t-PA antigen and its predictive value.Part of this paper was originally presented at the joint meetings of the 16th International Congress of the International Society of Fibrinolysis and Proteolysis (ISFP) and the 17th International Fibrinogen Workshop of the International Fibrinogen Research Society (IFRS) held in Munich, Germany, September 2002.

scholarly journals Analyzing Coronary Artery Disease in Patients with Low CAC Scores by 64-Slice MDCT

2012 ◽  
Vol 2012 ◽  
pp. 1-6 ◽  
Author(s):  
Nan-Han Lu ◽  
Lee-Ren Yeh ◽  
Tai-Been Chen ◽  
Yung-Hui Huang ◽  
Chung-Ming Kuo ◽  
...  
Keyword(s):  
Coronary Artery Disease ◽  
Logistic Regression ◽  
Coronary Artery ◽  
Meta Analysis ◽  
Prediction Method ◽  
Computed Tomographic Angiography ◽  
Health Examination ◽  
Physiological Variables ◽  
Proposed Model ◽  
Artery Disease

Purpose. Coronary artery calcification (CAC) scores are widely used to determine risk for Coronary Artery Disease (CAD). A CAC score does not have the diagnostic accuracy needed for CAD. This work uses a novel efficient approach to predict CAD in patients with low CAC scores.Materials and Methods. The study group comprised 86 subjects who underwent a screening health examination, including laboratory testing, CAC scanning, and cardiac angiography by 64-slice multidetector computed tomographic angiography. Eleven physiological variables and three personal parameters were investigated in proposed model. Logistic regression was applied to assess the sensitivity, specificity, and accuracy of when using individual variables and CAC score. Meta-analysis combined physiological and personal parameters by logistic regression.Results. The diagnostic sensitivity of the CAC score was 14.3% when the CAC score was ≤30. Sensitivity increased to 57.13% using the proposed model. The statistically significant variables, based on beta values andPvalues, were family history, LDL-c, blood pressure, HDL-c, age, triglyceride, and cholesterol.Conclusions. The CAC score has low negative predictive value for CAD. This work applied a novel prediction method that uses patient information, including physiological and society parameters. The proposed method increases the accuracy of CAC score for predicting CAD.

Sign in / Sign up

Export Citation Format

Share Document