Recently, random forest (RF) as a highly flexible machine learning algorithm has been applied to medicine, biology, machine learning, computer vision and other fields, and has shown good application performance. Nevertheless, the operation efficiency and identification accuracy of RF algorithm are actually affected by the number of the decision trees. A novel RF model, referred to as the extreme random forest (ERF), was proposed to improve the ability of feature extraction and reduce the computation burden. In the ERF method, the dimensionality of the high-dimensional data is randomly reduced through the random mapping matrix, and the classification performance after dimensionality reduction is improved. In this way, the sample dimension of the input RF is greatly reduced, which improves the operation efficiency of the RF. Both theoretical analysis and experiment tests have verified the superiority of the proposed method. In the experimental part, the present ERF method was compared with other peer method in terms of diagnostic performance and computational efficiency. The comparison results showed that the ERF method has more advantages both in diagnostic accuracy and computational efficiency. In addition to mechanical fault diagnosis, the proposed ERF can also be used in other machine learning fields.

Background: Many studies have shown that glycated hemoglobin (HbA1c) is associated with coronary artery disease (CAD). HbA1c was independently related to angiographic severity in Chinese patients with CAD after adjusting for other covariates. Some traditional cardiovascular drugs may have an impact on this relationship. Methods: This retrospective study enrolled a total of 572 CAD patients who underwent their coronary angiography and had their HbA1c levels measured at the Chinese Hospital. The complexity of the coronary artery lesions was evaluated using the Syntax score, and the subjects were divided into 4 inter quartiles according to HbA1c levels. Covariates included history of traditional cardiovascular drugs. Results:The average age of selected participants was 61.00 ± 9.15 years old, and about 54.72% of them were male. Result of fully adjusted linear regression showed that HbA1c was positively associated with Syntax score after adjusting confounders (β = 1.09, 95% CI: 0.27, 1.91, P = 0.0096). By interaction and stratified analyses, the interactions were observed based on our specification including with the medication history of statins and angiotensin receptor blockers (ARBs) (P values for interaction <0.05). Conclusion:In this study, we found a positive correlation between the HbA1c levels and the SYNTAX score among CAD individuals, and oral statins and ARBs medication could affect the correlation. Thus, HbA1c measurement could be used for the evaluation of the severity and complexity of coronary lesions among CAD patients.

BackgroundNumerous studies have demonstrated that the low-density lipoprotein cholesterol/high-density lipoprotein cholesterol (LDL-C/HDL-C) ratio can reflect the positive correlation index LDL-C and the negative index HDL-C of coronary artery disease (CAD) at the same time, which is increasingly considered as a novel marker to evaluate the risk of CAD. However, whether the short-term evaluation effect of the LDL-C/HDL-C ratio can be maintained during long-term follow-up is unclear. In addition, it is not clear whether the value of LDL-C/HDL-C ratio in the risk assessment of major adverse cardiac events (MACE) varies with different treatments. Our aim of the study was to investigate the link between LDL-C/HDL-C ratio and long-term risk of CAD and find out whether the LDL-C/HDL-C ratio could effectively evaluate the occurrence of MACE in CAD patients under different treatments. MethodsFrom May 2013 to November 2015, a total of 2409 patients who underwent coronary angiography (CAG) with or without revascularization therapy were enrolled in this study. They were divided into two groups based on the LDL-C/HDL-C ratio and three groups based on the treatments: medical therapy alone (MTA), percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG).ResultsIn total, 1784 patients (74.1%) were followed for health outcome and 625 patients (25.9%) experienced a MACE event. The median follow-up time was 4.27 years (1560 days). The patients with a higher LDL-C/HDL-C ratio (≥ 2.33) also had a significantly higher incidence of MACE (HR: 1.47, 95% CI: 1.25 to 1.72, p < 0.001). The cumulative incidence of rehospitalization for UA (HR: 1.53, 95% CI: 1.27 to 1.84, p < 0.001) and rehospitalization for HF (HR: 3.70, 95% CI: 1.22 to 22.25, p = 0.021) were significantly higher in high group than in low group. There were no significant differences in MI (HR: 1.25, 95% CI: 0.63 to 2.48, P = 0.521), TLR (HR: 0.98, 95% CI: 0.62 to 1.55, p = 0.947), Stroke (HR: 1.65, 95% CI: 0.64 to 4.25, p = 0.301) and 4-year all-cause death (HR: 1.45, 95% CI: 0.58 to 3.61, p = 0.423). Kaplan-Meier cumulative curve showed that patients with higher LDL-C/HDL-C ratio had a significantly lower MACE-free survival (p < 0.001). Multivariate Cox regression analysis demonstrated that LDL-C/HDL-C ratio (HR: 1.34, 95% CI: 1.14 to 1.60, p < 0.001) together with age, smoking, hypertension, diabetes mellitus, Syntax score and TG were independent predictors of 4-year MACE in the total CAD population (all p < 0.05). Further subgroup analysis showed that age, smoking, Syntax score, TG and LDL-C/HDL-C ratio were the independent predictors of MACE in MAT group (all p < 0.05); However, Syntax score and diabetes mellitus were the only independent predictor of MACE in PCI group and the CABG group, respectively (both p < 0.05). ConclusionsIn this study, we found that LDL-C/HDL-C ratio was an independent predictor of 4-year MACE in the total CAD population. The value of LDL-C/HDL-C ratio in assessing MACE risk varied among CAD patients with different treatments.