نتایج جستجو برای: روش mce
تعداد نتایج: 370568 فیلتر نتایج به سال:
BACKGROUND The neutrophil-to-lymphocyte (N/L) ratio has been associated with poor prognosis in patients with heart failure, but it has not been compared with N-terminal pro-brain natriuretic peptide (NT-proBNP) in elderly patients with chronic heart failure (CHF). We sought to make this comparison. METHODS A total of 1355 elderly patients with CHF were analyzed. A multivariate logistic regres...
Incremental Speaker Adaptation with Minimum Error Discriminative Training for Speaker Identification
Minimum Classification Error (MCE) has shown to be effective in improving the performance of a speaker identification system [1]. However, there are still problems to solve, such as the variability of the voice characteristics of a particular speaker through time. In this work, we analyze the degradation of a GMM-based textindependent speaker identification system when using test data recorded ...
Incremental speaker adaptation with minimum error discriminative training for speaker identification
Minimum Classification Error (MCE) has shown to be effective in improving the performance of a speaker identification system [1]. However, there are still problems to solve, such as the variability of the voice characteristics of a particular speaker through time. In this work, we analyze the degradation of a GMM-based textindependent speaker identification system when using test data recorded ...
Objective: To assess the role of intravenous myocardial contrast echocardiography (MCE) in predicting functional recovery and regional or global left ventricular (LV) remodelling after acute myocardial infarction (AMI) compared with low dose dobutamine stress echocardiography (LDSE). Methods: 21 patients with anterior AMI and successful primary angioplasty underwent MCE and LDSE during the suba...
AIM To investigate the effect of abciximab on microvascular integrity and left ventricular (LV) functional recovery in patients with acute myocardial infarction (MI) treated by primary coronary angioplasty (PTCA). METHODS AND RESULTS Thirty-one patients (27 males; age 39-76 years) with first, acute MI (<6 h after onset) were randomized to receive either abciximab+primary PTCA (n=17) or primar...
OBJECTIVE The aim of this study was to evaluate the value of real-time myocardial contrast echocardiography (RT-MCE) for detecting coronary microcirculation (CM) function in coronary artery disease (CAD) patients. METHODS Sixty-five consecutive patients were divided into CAD (n=52) and no-CAD (n=13) groups using coronary angiography (CAG). All patients underwent RT-MCE at rest and CAG within ...
AIMS We hypothesized that molecular imaging of endothelial P-selectin expression with targeted myocardial contrast echocardiography (MCE) could identify recently ischaemic myocardium without infarction. METHODS AND RESULTS The microvascular behaviour of P-selectin-targeted (MB(p)) and control (MB(c)) microbubbles was assessed by intravital microscopy of the cremaster muscle in mice. Targeted ...
U p to 30% of patients with acute myocardial infarction (AMI) treated with reperfusion and with an apparently successful result have microvascular dysfunction. Multiple techniques are available for the diagnosis of this condition, but little information is available comparing their ability to predict ventricular recovery after AMI. The objective of our study was to compare the accuracy of myoca...
Discriminative training of hidden Markov models (HMMs) using segmental minimum classi cation error (MCE) training has been shown to work extremely well for certain speech recognition applications. It is, however, somewhat prone to overspecialization. This study investigates various techniques which improve performance and generalization of the MCE algorithm. Improvements of up to 7% in relative...
This article shows that the Minimum Classification Error (MCE) criterion function commonly used for discriminative design of speech recognition systems is equivalent to a Parzen window based estimate of the theoretical Bayes classification risk. In this analysis, each training token is mapped to the center of a Parzen kernel in the domain of a suitably defined random variable. The kernels are s...
نمودار تعداد نتایج جستجو در هر سال
با کلیک روی نمودار نتایج را به سال انتشار فیلتر کنید