[Interpretation of experimental data in the presence of accidental perturbations: characterization of the bladder].

Some tests are subject to the influence of a special kind of "irreducible" effects which do not satisfy the Central Limit Theorem of statistics, even though they show no systematic character or tendency. These effects especially present when physiological experiments are concerned "in vivo". In such the lack of specific methods to reduce these effects brings about the necessity of special case in minimizing the loss of information. Contained in the experimental data as well as the accumulation of useless information. Recent work has suggested the extensive use of several indexes derived from combination of available variables and their respective mathematic correlation. The information obtained by following this methodology has the character of "working hypothesis" so it needs many favorable confirmation to assess its reliability. It is convenient to analyze a large number of relations in order to emphasize the concordant effects and discord the others. In order to obtain some preliminary indication of the reliability of this procedure we have analyzed some CMG curves in prostatic and non-prostatic subject. The final slope of the curves (urge point) t = dP/dVr has been adapted as an index of the "elasticity" of the bladder; it was compared with the filling volume, the muscular volume and their ratio, i.e. with a geometric index of the bladder shape. The correlation show that the index t can, to some extent significantly represent the mechanical behaviour of the bladder in its physiological condition or when there is a partial obstruction of urine deflection.(ABSTRACT TRUNCATED AT 250 WORDS)