Hitherto unpublished technique for prospective diagnostics of deleterious human diseases based on the generalization of the chromatin organization in human cell nuclei is introduced. Its principle is in the interaction with a model derived from image data, which are collected by the fully automated high resolution cytometer. The most significant topological descriptors of the cell nuclei are selected and processed in the inductive learning system. In its training phase the cytological properties of the analyzed cell as well as all the external stimuli are known and used for the proper adaptation. Then the resultant structure serves as a classifier categorizing unknown biological material in accordance with internal structure or parameters. Format of the input data and possible approaches to the problem formalization are presented and documented with several examples. Lastly, the experimental results are discussed and future research directions are suggested.