ABSTRACT
Background: Medical/health researchers depend on data evidence for knowledge discovery. At times, data analysis to capture the data evidence is overwhelming and the process becomes too tedious to give up the attempt. A prudent thing to do is to seek out a simpler visual approach to obtain insights. One visual approach is devised in this article to understand what the data are really revealing to either get an insight first or then confirm what is intuitively configured by the medical concepts. This visual approach is geometric concepts based. In specific, triangle is employed in this new and novel approach. Methods: A successful treatment of any illness is a consequence of knowledge build-up arising from data mining about the never, once, or repeated episode of a disease incidence in a patient. This article investigates and illustrates a novel and pioneering geometric approach, especially based on the properties of triangle, to extract hidden evidence in the data. New probabilistic expressions are derived utilizing trigonometric relations among the corner points of a triangle. The conceptual contents of this article are versatile enough for different medical/health data analysis. Results: For illustration here, the medical binomial data in Hopper et al. (Genetic Epidemiology, 1990) on the occurrence of asthma or hay fever among the four groups: (1) monozygotic females (MZF), (2) monozygotic males (MZM), (3) di-zygotic females (DZF), and (4) di-zygotic males (DZM) are considered and triangularly interpreted. The results indicate that the angle in the vertex representing one episode is the largest compared to the other two angles in the vertices representing never or repeated episode of an illness among a random sample of twins from these four groups with respect to getting asthma or hay fever. This geometric finding implies that the event of never and the event of repeated incidence of the illness have farthest Euclidean distance in probability sense. In other words, the never and repeated incidences are not in close proximity as probable. Conclusions: This geometric view of this article is versatile enough to be useful in other research studies in drug assessment, clinical trial outcomes, business, marketing, finance, economics, engineering and public health whether the data are Poisson or inverse binomial type as well.
ABSTRACT
Biovolume calculation of micro algae is commonly used for the assessment in relative abundance (as biomass or carbon) of coexistent algae that vary in shape and/or size. Periphyton samples were obtained within a more general program, in fishery tanks and streams situated in the upper basin region of Tibagi River, nearby Londrina, Paraná state, south of Brazil. The present work proposes new algorithms for the determination of algal species volume. Results showed that Gyrosigma scalproides (Rabenhorst) Cleve 1894 and Tetraedron gracile (Reinsch) Hansgirg 1889 were exceptions inside their genus and the present new formula for biovolume and area gives more adequate results. In addition, new formulas applied for Phacus longicauda (Ehrenberg) Dujardin 1841 and Kirchneriella lunaris (Kirchner) Möbius 1894 also showed much more accurate results than in the previous works. The difference in the calculations of biovolume between the present method and others was not small and could not be neglected.
ABSTRACT
PURPOSE: To assess wall motion of the left ventricle in the various forms of congenital heart disease and to select appropriate methods by which to gauge left ventricular wall motion. METHODS: We evaluated left ventricular function, including volume, global ejection fraction, and regional wall motion(Centerline method, Local Ejection Fraction, Curved Perimeter method and Radial method). The evaluation was done on review mode of 30-degree elongated right anterior oblique projection of left ventriculogram by using cardiac analysis program of the DXC Hiline Digital System(GER, USA) in 53 pediatric patients undergoing diagnostic(or therapeutic) cardiac catheterization from June 1995 to July 1996 at Kyung-pook National University Hospital. RESULTS: Cardiac indices calculated by Dodge's and Simpson's methods were 5.76+/-2.13 L/min/ m2, 5.86+/-2.31 L/min/m2 respectively with significant correlation to each other(r=0.99, P<0.01). Global ejection fractions using both methods were 69.81+/-8.33%, and 69.69+/-8.20%, respectively (r=0.99, P<0.01). After dividing the left ventricular outline contour into five regions such as posterobasal, diaphragmatic, apical, anterolateral and anterobasal portions, all four models could ascertain regional wall motion abnormalities well(P<0.01). And among them, the Centerline method and the Curved Perimeter method correlated well in assessing wall motion(for each portion r=0.67, 0.84, 0.94, 0.91 and 0.61 respectively, P<0.01). CONCLUSION: In conclusion, we confirmed regional wall motion abnormalities in the various forms of congenital heart disease in spite of normal global ejection fractions, showing that the Centerline and Curved Perimeter methods are preferable methods in the assessment of wall motion in patiens with congenital heart disease.