Background: To investigate the clinical value of acoustic cardiography in the diagnosis of coronary artery disease (CAD) and post-percutaneous coronary intervention (PCI) early asymptomatic left ventricular systolic dysfunction. Methods: Inpatients in the department of cardiology were included in the research (n = 315); including 180 patients with angina pectoris and 135 patients with acute anterior wall myocardial infarction after emergency PCI did not present with signs and symptoms of heart failure. Color Doppler echocardiography, brain natriuretic peptide, acoustic cardiography examination were performed. The patients were divided into four groups: non-CAD group (n = 60), CAD group (n = 120), MI-REF group (EF% < 50%, n = 75), and MI-NEF group (EF% ≥ 50%, n = 60).Results: Acoustic cardiography parameters EMATc, systolic dysfunction index, S3 strength and S4 strength in the MI-REF group were higher than those in MI-NEF group (p < .05), and the MI-NEF group was higher than CAD group (p < .05). S3 strength (area under the curve [AUC] 0.67, 95% CI 0.585-0.755, p < .001) and S4 strength (AUC 0.617, 95% CI 0.536-0.698, p = .011) are useful in the diagnosis of CAD. S3 strength (AUC 0.942, 95% CI 0.807-0.978, p < .001) was superior to other indicators in the diagnosis of early left ventricular systolic dysfunction after myocardial infarction. Conclusion:S4 combined with ST-T standard change can improve the diagnosis of CAD. Acoustic cardiography can be used as a non-invasive, rapid, effective, and simple method for the diagnosis of asymptomatic left ventricular systolic dysfunction in the early stage after myocardial infarction.

The fault diagnosis for power transformer plays an important role in improving the safety and reliability for an electrical network. Dissolved gas analysis (DGA) is a basic method to diagnose the fault of power transformer. Considering the disadvantages in DGA using fuzzy c-means (FCM) clustering algorithm, a hybrid method based on particle swarm optimization (PSO) to solve the FCM model is presented. In the new algorithm, the PSO’s search space is the vector space after straightening the membership matrix in the FCM. With the results of experiments on real DGA data, it shows our approach can improve the clustering performance for the transformer fault diagnosis.

Risk management and its accurate analysis are very important for project management. RBF and MLP Neural Network Model are common methods of risk management and analysis, which are not accurate enough. In this paper a new method based on LS-SVM is introduced. Analytical models of risky projects are investigated and function approximation results are compared. Experimental results show that the regression analysis of risk based on LS-SVM method has higher prediction accuracy and better generalization ability.