ABSTRACT
This paper presents a novel method for discrimination between innocent and pathological murmurs using the growing time support vector machine (GTSVM). The proposed method is tailored for characterizing innocent murmurs (IM) by putting more emphasis on the early parts of the signal as IMs are often heard in early systolic phase. Individuals with mild to severe aortic stenosis (AS) and IM are the two groups subjected to analysis, taking the normal individuals with no murmur (NM) as the control group. The AS is selected due to the similarity of its murmur to IM, particularly in mild cases. To investigate the effect of the growing time windows, the performance of the GTSVM is compared to that of a conventional support vector machine (SVM), using repeated random sub-sampling method. The mean value of the classification rate/sensitivity is found to be 88%/86% for the GTSVM and 84%/83% for the SVM. The statistical evaluations show that the GTSVM significantly improves performance of the classification as compared to the SVM.
Subject(s)
Heart Murmurs/classification , Phonocardiography/methods , Support Vector Machine , Adolescent , Aged , Aged, 80 and over , Aortic Valve Stenosis/classification , Aortic Valve Stenosis/physiopathology , Child , Child, Preschool , Databases, Factual , Heart Murmurs/physiopathology , Humans , Infant , Middle Aged , Sensitivity and Specificity , Severity of Illness Index , Wavelet AnalysisABSTRACT
BACKGROUND: A novel polymer-shelled contrast agent (CA) with multimodal and target-specific potential was developed recently. To determine its ultrasonic diagnostic features, we evaluated the endocardial border delineation as visualized in a porcine model and the concomitant effect on physiological variables. METHODS: Three doses of the novel polymer-shelled CA (1.5 ml, 3 ml, and 5 ml [5 × 10(8) microbubbles (MBs)/ml]) and the commercially available CA SonoVue (1.5 ml [2-5 × 10(8) MBs/ml]) were used. Visual evaluations of ultrasound images of the left ventricle were independently performed by three observers who graded each segment in a 6-segment model as either 0 = not visible, 1 = weakly visible, or 2 = visible. Moreover, the duration of clinically useful contrast enhancement and the left ventricular opacification were determined. During anesthesia, oxygen saturation, heart rate, and arterial pressure were sampled every minute and the effect of injection of CA on these physiological variables was evaluated. RESULTS: The highest dose of the polymer-shelled CA gave results comparable to SonoVue. Thus, no significant difference in the overall segment score distribution (2-47-95 vs. 1-39-104), time for clinically sufficient contrast enhancement (20-40 s for both) and left ventricular overall opacification was found. In contrast, when comparing the endocardial border delineation capacity for different regions SonoVue showed significantly higher segment scores for base and mid, except for the mid region when injecting 1.5 ml of the polymer-shelled CA. Neither high nor low doses of the polymer-shelled CA significantly affected the investigated physiological variables. CONCLUSIONS: This study demonstrated that the novel polymer-shelled CA can be used in contrast-enhanced diagnostic imaging without influence on major physiological variables.
