Processing Technique of Weighted Data to Improve Assessment of Latent Information

There is growing interest in revealing hidden information in datasets. Data processing techniques allow both to reduce the economic costs of obtaining the necessary information and to improve the quality of decision making based on this data. The focus of this paper is latent quantitative data extracted on weighted images and high quality decision to separate of normal or diseased states subjects. Complex processing technique allows to identifying subjects' states and can help make medical diagnostics more deliberate and increase price-performance. This approach increases the chances to choose the most satisfying result within alternative possible. The data base of research is tomograms of patients with Alzheimer's disease that have compared with data of probably normal and of mild cognitive impairment patients. A prior study suggested that artificial presented indicator named iron index has in the basal ganglia of patients with Alzheimer disease higher level then one of control subjects. In the current study the factor analysis was used to improve separation of large numbers of observed variables from image data set that are thought to reflect a smaller number of underlying disease states.