
Payment regulatory mechanism as a source of wage increases in healthcare
Grim, Jiří
The principle of health insurance presupposes that the patient will contact a doctor who will provide him / her with professional help, whereby the doctor, medical expenses and additional examinations are paid by the health insurance company. The result is a spontaneous increase in health care costs wellknown in the nineties. It is clear that there is no negative feedback in the system where the healthcare provided must be made by doctors in contact with patients and its costs are being covered by health insurance companies. As a result of this gross systemic error, there is a continuing pressure to increase healthcare spending and imminent insolvency forces the health insurers to introduce regulatory measures to curb the cost increase.

 

Feasibility Study of an Interactive Medical Diagnostic Wikipedia
Grim, Jiří
Considering different application possibilities of product distribution mixtures we have proposed three formal tools in the last years, which can be used to accumulate decisionmaking knowhow from particular diagnostic cases. First, we have developed a structural mixture model to estimate multidimensional probability distributions from incomplete and possibly weighted data vectors. Second, we have shown that the estimated product mixture can be used as a knowledge base for the Probabilistic Expert System (PES) to infer conclusions from definite or even uncertain input information. Finally we have shown that, by using product mixtures, we can exactly optimize sequential decisionmaking by means of the Shannon formula of conditional informativity. We combine the above statistical tools in the framework of an interactive openaccess medical diagnostic system with automatic accumulation of decisionmaking knowledge.


Approximating Probability Densities by Mixtures of Gaussian Dependence Trees
Grim, Jiří
Considering the probabilistic approach to practical problems we are increasingly confronted with the need to estimate unknown multivariate probability density functions from large highdimensional databases produced by electronic devices. The underlying densities are usually strongly multimodal and therefore mixtures of unimodal density functions suggest themselves as a suitable approximation tool. In this respect the product mixture models are preferable because they can be efficiently estimated from data by means of EM algorithm and have some advantageous properties. However, in some cases the simplicity of product components could appear too restrictive and a natural idea is to use a more complex mixture of dependencetree densities. The dependence tree densities can explicitly describe the statistical relationships between pairs of variables at the level of individual components and therefore the approximation power of the resulting mixture may essentially increase.

 
 

Evaluation of Screening Mammograms by Local Structural Mixture Models
Grim, Jiří ; Lee, G. L.
We consider the recently proposed evaluation of screening mammograms by local statistical models. The model is defined as a joint probability density of inside grey levels of a suitably chosen search window. We approximate the model density by a mixture of Gaussian densities. Having estimated the mixture parameters we calculate at all window positions the corresponding loglikelihood values which can be displayed as grey levels at the respective window centers. The resulting loglikelihood image closely correlates with the original mammogram and emphasizes the structural details. In this paper we try to enhance the loglikelihood images by using structural mixture model capable of suppressing the influence of noisy variables.

 

Fast DependencyAware Feature Selection in VeryHighDimensional Pattern Recognition Problems
Somol, Petr ; Grim, Jiří
The paper addresses the problem of making dependencyaware feature selection feasible in pattern recognition problems of very high dimensionality. The idea of individually best ranking is generalized to evaluate the contextual quality of each feature in a series of randomly generated feature subsets. Each random subset is evaluated by a criterion function of arbitrary choice (permitting functions of high complexity). Eventually, the novel dependencyaware feature rank is computed, expressing the average benefit of including a feature into feature subsets. The method is efficient and generalizes well especially in veryhighdimensional problems, where traditional contextaware feature selection methods fail due to prohibitive computational complexity or to overfitting. The method is shown well capable of overperforming the commonly applied individual ranking which ignores important contextual information contained in data.


Diagnostické vyhodnocování screeningových mamogramů pomocí lokálních texturních modelů
Grim, Jiří ; Somol, Petr
We propose statistically based preprocessing of screening mammograms with the aim to emphasize suspicious areas. We estimate the local statistical texture model of a single mammogram in the form of multivariate Gaussian mixture. The probability density is estimated from the data obtained by pixelwise scanning of the mammogram with the search window. In the second phase, we evaluate the estimated density at each position of the window and display the corresponding loglikelihood value as a gray level at the window center. Light gray levels correspond to the typical parts of the image and the dark values reflect unusual places. The resulting loglikelihood image exactly correlates with the structural details of the original mammogram, emphasizes locations of similar properties by contour lines and may provide additional information to facilitate diagnostic interpretation.
