Prof.  Riad Taha  Al-Kasasbeh
The University of Jordan,  Jordan

Title: CLASSIFIERS OF THE FUNCTIONAL STATE OF LIVING SYSTEMS BASED ON VOIGT MODEL OPTIMIZATION

Abstract:

This work introduces a bioimpedance spectroscopy method that can be used for multiple purposes; this method can help in the development of objective and realistically accessible criteria for disease assessment. It can assess the severity and prognosis of diseases, in addition to evaluating the effectiveness of treatment methods. Furthermore, the proposed method can be utilized to develop criteria for conservative therapy options and surgical interventions in severe patients.

The proposed method employs a recurrent Voigt model to represent biomaterial segment impedance. For each biomaterial segment model, a Cole plot is drawn in a given frequency range. In order to determine the parameters of each model, a recurrent procedure is performed, which involves solving multiple systems of nonlinear equations. The solution starts with one link of the Voight models, and then moves forward at each iteration step, until the the approximation error of the Cole experimental plot reaches a predetermined value. A recurrent algorithm for optimizing biomaterial Voight models has been developed. The algorithm is based on the results of bioimpedance analyses, which are presented in the form of impedance diagrams. The algorithm is distinguished by the use of a recurrent procedure for comparing the biomaterial impedance models obtained on the basis of sequentially connected links of the Voight model, as well as by the method of forming frequency pools. In the optimization process, based on a recurrent procedure, the optimal number of links in the Voight model and frequency pools is established. Based on the analysis of the integral error, the optimal number of model links is selected, and the optimal frequency pool corresponding to this number is determined. The Voight model parameters allow the formation of descriptors for multimodal classifiers of the functional state of living systems and parameters of the model links.

In order to confirm the validity of the proposed classifier, a group of patients diagnosed with pneumonia has been tested. To obtain bioimpedance analysis data, an electrode belt was placed on the chest of each patient, and impedance diagrams corresponding to a certain combination of electrodes were determined. The quality indicators of various classifier models reached 78% and were above 62%.

Biography:

Professor Riad Taha Al-Kasasbeh received his MS in Engineering Science and Ph.D. in Controlling Biological and Electronic Equipment, and his PhD from Saint-Petersburg Electrotechnical University ETU “LETI”, He was a postdoctoral fellow at Konstanz University (Germany). He is currently a Professor at the University of Jordan. He is a member of professional organizations and co-author of over 110 papers (Editions: Springer, IEEE, Taylor & Francis, IASTED, Inderscience, Elsevier, etc.). He is a Visiting Professorship at several Universities including Philadelphia University, Konstanz University (Germany)  (HTWG), and Karaganda State Industrial University ( Kazakhstan), "Moscow Power Engineering Institute" (MPEI). He was a Research Fellow of DFG at the HTWG. His research interests include biomedical instrumentation , biophysics, acupuncture, fuzzy logic decision-making, and medical and ecology information systems, Ergonomics. He presented his work at numerous international conferences as a keynote speaker. In addition. He has developed novel clinical methods for the prediction and early diagnosis of heart , stomach  diseases, also he has determined features for determining the level of psycho-emotional tension in the man-machine system operators using bio-active points based on fuzzy logic rules. As well as the projects manager in research and development projects that performs scientific research for the Industry and medical departments and organizations local and international .