The submission should be no more than 2 A4 pages of text in a word file (1800 words) or up to three pages if it includes diagrams, tables and references, with the extra space devoted to these and any captions. Font size should be no smaller than 10 pt Times New Roman with 2.5cm margins. All pages should be numbered and diagrams and tables should be positioned in their proper place in relation to the text, ensuring that the total contribution does not exceed 3 pages. Authors invited to prepare full manuscripts for submission to Interface Focus will also need to adhere to a further schedule for submission and review; final copy will be required by 15th February 2011.
submitted as extended_abstract.odt
Jiří Kofránek, Tomáš Kulhánek, Marek Mateják, Pavol Privitzer, Jan Šilar, Martin Tribula
The computational models of human physiology was being developed at our Laboratory of Biocybernetics and Computer Aided Teaching  using Matlab/Simulink. Our current models cover most important parts of human physiology and reuses published models and schemas e.g. from circulatory physiology . The web based educational simulators utilizing these models are being used during presentation and lessons of the topic Physiology at First Faculty of Medicine, Charles University and we moved from casual modeling in Matlab/Simulink to non-casual modeling in Modelica . Cooperating with the association CESNET we connected to a pilot grid infrastructure which joins several physical servers spread among different locations connected via high speed academic network CESNET2 to a computational and data grid.
With cooperation of partners from clinics and laboratory we are researching methods that will improve computational abilities to support comparison of measurements with a computed physiological models - so called identification of physiological systems. The computational models have set of parameters, some of them corresponds to the measurable quantities. These can be obtained during pacient examination. However, some of the parameters are not measurable by any techniques or it's hard to measure them, but usually varies around known normal physiological values. The identification of a parameter is the process to determine values of such model's parameter, which cannot be measured or guessed. The simulated model behavior based on identified parameter values is compared with the experimentally measured quantities. Identification of large number of parameters implicates searching them in multidimensional space which might become computation demanding task. Thus the methods to parallelize the computation is being developed and integrated into grid computing middleware and the mentioned virtual grid infrastructure.
For identification of parameters we tested these optimization methods: simplex method , conjugate gradient method , genetic algorithms (CMA-ES), simulated annealing, multistart methods. We also tested these inversion methods: ANNIT  and isometric  methods. In the first stage we implemented these identification methods in MATLAB. Our models implemented in Modelica language are compiled into computation libraries in .NET. We deployed .NET computation libraries as a remote services which are called by identification methods via TCP/IP connection. The preliminary results shows that the optimization methods might be deployed easily in distributed environment and executed in parallel computing system.
Integration of computation of big and complex physiological models is possible into distributed computing. The non-casual modeling in Modelica provides flexible system to modify model and behavior during computation or among computation cycles. However, the proprietary communication between .NET computation libraries and optimization methods brings some unexpected difficulties among several platforms and data type interpretation. Thus there is plan to develop and utilize some common techniques or standard e.g. web services or REST services over HTTP. We plan to integrate a common system to distribute computational libraries among several clients, e.g. BOINC 
 Physiolibrary for Matlab/Simulink, (1997-2007), web: http://patf-biokyb.lf1.cuni.cz/wiki/projekty/physiolibrary
 S. R. Abram, B. L. Hodnett, R. L. Summers, T. G. Coleman, R. L. Hester: Quantitative circulatory physiology: an integrative mathematical model of human physiology for medical education. Adv. Physiol. Educ., 31, 202-210, (2007).
 Stodulka Petr, Privitzer Pavol, Kofránek Jiří: Web-based Educational Simulators for Teaching Pathological Physiology , EISTA 2009, IMSCI 2009, Orlando, FL, USA, p. 100-102, 2009
 Kulhanek Tomas, Sarek Milan: Processing of medical information in virtual distributed environment. In EATIS 09 Contribution Proceedings. Euro American Conference on Telematics & Information Systems, Prague 3-5 June 2009, Wirelesscom s.r.o., 2009, ISBN #978-1-60558-398-3
 J. A. Nelder and R. Mead, „A simplex method for function minimization“, Computer Journal, 1965, vol 7, pp 308–313
 Hestenes, Magnus R.; Stiefel, Eduard (December 1952). Methods of Conjugate Gradients for Solving Linear Systems, web: http://nvl.nist.gov/pub/nistpubs/jres/049/6/V49.N06.A08.pdf
 B. Růžek (2008), Artificial Neural Network Inversion Tool, web: http://www.ig.cas.cz/userdata/files/personal-pages/b-ruzek/ANNO/ANNIT_EN.pdf
 Málek J., Růžek B. and Kolář P. (2007), Isometric method: Efficient tool for solving non-linear inverse problems. Stud. Geophys. Geod., 51, 469-499. web: http://www.ig.cas.cz/userdata/files/personal-pages/b-ruzek/ANNO/IM_final.pdf
 BOINC, Open-source software for volunteer computing and grid computing. http://boinc.berkeley.edu/