Search

Numerous data mining models have been proposed to construct computer-aided medical expert systems. Bayesian network classifiers (BNCs) are more distinct and understandable than other models. To graphically describe the dependency relationships among clinical variables for thyroid disease diagnosis and ensure the rationality of the diagnosis results, the proposed k-dependence causal forest (KCF) model generates a series of submodels in the framework of maximum spanning tree (MST) and demonstrates stronger dependence representation. Friedman test on 12 UCI datasets shows that KCF has classification accuracy advantage over the other state-of-the-art BNCs, such as Naive Bayes, tree augmented Naive Bayes, and k-dependence Bayesian classifier. Our extensive experimental comparison on 4 medical datasets also proves the feasibility and effectiveness of KCF in terms of sensitivity and specificity. [ABSTRACT FROM AUTHOR]

Gastric intestinal metaplasia (GIM) is a precancerous lesion of the stomach. The detection of GIM using conventional white-light endoscopy (WLE) is difficult. In this study, we determined whether acetic acid-enhanced narrow band imaging (AA-NBI) improves the detection of GIM. A consecutive cohort of 132 individuals aged 40 years or older was subjected to upper gastrointestinal endoscopy using WLE, NBI and AA-NBI. The ability of the three methods to diagnose GIM in patients was compared. Histological assessment (per-patient and per-biopsy) was used for the accuracy assessment. Sixty-six (50.0%) out of the 132 individuals examined were found to have GIM, of which 44 (66.7%) were diagnosed correctly by NBI (sensitivity 66.7% and specificity 68.2%) and 58 (87.9%) were correctly identified by AA–NBI (sensitivity 87.9% and specificity 68.2%), as compared to only 22 (33.3%) by WLE (sensitivity 33.3% and specificity 28.8%). Therefore, the sensitivity of AA–NBI in the diagnosis of GIM was significantly higher than NBI (p<0.05) and WLE (p < 0.001). Our study indicates that AA-NBI can improve the accuracy of endoscopy-targeted biopsies for GIM. [ABSTRACT FROM AUTHOR]

The aim of this study was to evaluate the relationship between enhanced intensity of contrast enhanced ultrasound and microvessel density of aortic atherosclerotic plaque in rabbit model. The abdominal aortas of thirty-six male New Zealand rabbits were damaged by balloon expansion and the animals were then fed a high fat diet for 12 weeks. Twenty-seven plaques on the near aortic wall were detected using conventional ultrasound examination. The maximum thickness of each plaque was recorded. CEUS was performed on these 27 plaques and the time-intensity curves (TICs) were analyzed offline. Using the quantitative ACQ software, features such as the arrival time (AT), time to peak (TTP), baseline intensity (BI), peak intensity (PI) and enhanced intensity (EI) (EI = PI-BI) were assessed. Inter- and intra-observer agreement of EI were assessed using the Bland-Altman test. After CEUS examination, the rabbits were sacrificed for pathological examination and CD34 monoclonal antibody immunohistochemical detection. Microvessel density (MVD) was counted under the microscope. The relationship between indexes of CEUS and the level of MVD was analyzed. There was a good positive linear correlation between EI and MVD (γ = 0. 854, P<0. 001), the intraclass correlations for inter- and intra-observer agreement for EI were 0.73 and 0.82 respectively, suggesting that EI may be act as a useful index for plaque risk stratification in animal models. [ABSTRACT FROM AUTHOR]