The management of non-linear and grey clinical data, so frequent in psychiatry, needs the development of new statistical tools, able to give value to subjectivity and complexity of the observed changes.

A comparison between soft computing models, psychodynamic theory and clinical observations in the experience of psychiatric rehabilitation.

To test the use of some dynamic fuzzy algorithms (like clustering, ellipsoidal rules, least square method) in order to improve data analysis in the field of clinical research in psychiatry.

The role of slow, evolutionary, multifactorial, minute movements and their potential significance become readable, especially referred to the subjective observations of the clinical operators.

In this work, that it's necessary to consider only a first step, seems that Dynamic Fuzzy Functions may have a future role in the development of new assessments of clinical data in psychiatry.