[:en]The Project Archimedes ΙΙ is a re-enforcement program for Crete’s Technological Educational Institute is co-funded by the European Social Fund and National Resources – EPEAEK ΙΙ . The subproject is about Artificial Intelligence Applications in Software Technology and Medicine. The goal of this project is to develop software systems that rely on knowledge bases. There will be studied and developed advanced techniques of Artificial intelligence for de development of the systems.

 

Organizations

 

There are methods of developing and maintaining software that produce programs which are the design decisions used in their manufacture. Such programs or projects of programs can easily be maintained and used to prove properties but are not effective. In this project we will study and implement techniques that improve the effectiveness of such programs. In addition, we will study the issue of verification software and are aiming to build a verifier logic programs which are semiautomatic.

Also, based on techniques of Artificial Intelligence, will study parallel and distributed data structures and algorithms for the certification of systems and networks on chip. The libraries will develop tools to extend existing certification as SCV (SystemC Verification Library) and Testbuilder based on standard modeling SystemC2.0.

Finally, we intend to develop a decision support system which aims to assist in the classification of seizures that occur in childhood, according to the international classification of seizures. It will develop a database which will record incidents epilepsy (patient file with all relevant data). The implementation of both subsystems and control of these real working environment will help us to draw conclusions about the usefulness of similar systems in diagnostic procedures neurologist for the disease of epilepsy. The Project Archimedes ΙΙ is a re-enforcement program for Crete’s Technological Educational Institute is co-funded by the European Social Fund and National Resources – EPEAEK ΙΙ . The subproject is about Artificial Intelligence Applications in Software Technology and Medicine. The goal of this project is to develop software systems that rely on knowledge bases. There will be studied and developed advanced techniques of Artificial intelligence for de development of the systems.

 

Objectives

The objective of the project is about the institutionalization of Crete’s T.E.I. laboratory to promote scientific research in important fields such as Artificial Intelligence and Software Technology and especially in the fields of Logic Programming, On chip System and Network Methodology and Design and Medical Applications in cooperation with worldwide known Universities, Research centers and Companies. Al the activities of the subproject are about the abovementioned fields. We hope that this subproject will help in the research development and particularly to the best research presence, and also to the institutionalization and improvement of the T.E.I. of Crete “Artificial Intelligence and Software Technology” laboratory infrastructure. We feel that cooperation with top Universities and especially Companies will uphold Applied Technological Research of Crete’s T.E.I. .

 

Development

There are methodologies of development and preserve software that construct programs, which represent designing decisions used for their construction. Such programs or program designs, can easily be preserved and used for proving their properties but they are not effective. In this subproject we will study and develop techniques that improve the efficiency of these programs. Software verification is a quite difficult matter with a lot of open issues. We will study software verification matter and we aim to construct a semi-automatic verifier for logic programs.

Also, based on techniques of artificial Intelligence, we will study parallel and distributed data structures and algorithms, for on chip system and network verification. The libraries that will be developed will expand the existing verification tools, like SCV (SystemC Verification Library) and Testbuilder that are based on modeling pattern SystemC2.0.

In addition, methodologies that are developed in the area of Artificial Intelligence, give us the ability to organize medical knowledge, and represent it in a computational system, valorize it over the interpretation of clinical, demographic and laboratory data, which are helpful in the process of medical diagnosis and in the extraction of correct results. We intent to develop a decision support system whose target is to offer help at the classification of epileptic seizures that appear in childhood, according to the international classification for epileptic seizures as proposed by International Union against Epilepsy. At the same time we will develop a data base in which epilepsy incidents will be registered (patients record with clinical, laboratory and demographic data). The implementation of the two above subsystems and their function control in a real work environment , will help us extract conclusions about the usability of the respective systems in diagnostic procedures of neurologists for the condition of epilepsy.
111

Results

  1. A)The construction of an automatic or [semi-automatic] tool which will convert the very structured programs in respectively compacter (smaller code) and more effective programs. 2) Growth of one dialogic (verifier) that will prove the correctness of manufactured programs. Moreover, will be studied the relation which could exist to the structure of manufactured programs and in the structure of proof of its correctness . The ineffectiveness of programs that is manufactured with method [Marakakis94, Marakakis97] can be owed in their very structured form. The inefficiency of these programs can be corrected by applying rules of transformation of logic programs as mentioned before. The objective of these transformations is to remove from the programs the designing level, and to decrease their size and make them more effective.
  2. B) We will study friendly to the user environment of verification, designing and implementing a parallel or distributed tool of static and dynamic analysis, that gives possibilities of verification the system properties based to the user demands. This tool, based on graph theory and combinational methods (combinatorics), could also confirm conditions of real time in complex, probably asynchronous models of systems, with the use of feasible tables (reachability matrices), or with calculations of reliability.
  3. C) Construction of a system for children’s epilepsy diagnosis support , that will help the doctor to classify an incident of epileptic crisis in concrete category of international classification [ILAE81, ILAE89] and accordingly to contribute considerably in the diagnostics process, that usually follows the doctor in cases of epilepsy. At the same time the possibility of search of similar incidents from the base of data (patient’s filel) will give him the chance examine proportional situations, studying corresponding diagnostic processes, and therapeutic methodologies that were applied.

 

Impact

  1. A)The software that will be constructed as well as the method of construction of logic programs [Marakakis94, Marakakis97] will be used in courses with regard to the development of software (Programming II, Technology of Software). Moreover, will be developed advanced techniques of Artificial Intelligence. They will be educationally used in the course Artificial Intelligence. We aim to come in contact with companies of software development that will develop more the results of this activity.

For the Program Transformation Software and for the Software Verifier interests the computer science company “DataCreta” which develops and supports commercial software, Heraklion, Crete. Cf. attached letter of event of interest of company “DataCreta”.

  1. B)Most errors that are discovered at the duration of simulation could be discovered earlier, with smaller cost using high level techniques of analysis and simulation.. the programming environment that will be constructed will be also used in corresponding courses of choice of TEI in order to valuate of combinational inventive techniques (combinatorics, heuristics) and improved the efficiency of the system with new, more efficient algorithms. Also we hope in collaboration with the industry for its suitable exploitation.

For the system Verification for On chip systems and networks of communication , there is interest from the laboratory of ST Microelectronics, “AST Advanced System Technology”, Grenoble, France. This appears from the publications and the presence of company ST Microelectronics as main collaborator in open code software for networks of communications in chip (http://www.occn.sourceforge.net) which it is supported til today by the main researcher Dr M. Grammatikakis.

  1. C)The increasing needs to give qualitative services of health, but also the rapid rythms of medical knowledge renewal, require from the doctor to dedicates a lot of time in every case he examines. Into practice however most medical decisions should be taken in limited time and usually they are based on knowledge that exists at that moment in the brain of doctor. Thus in their majority the medical decisions are results of hasty decisions , because the doctor infrequent has the chance, due to lack of time mainly, to be addressed in some handbook/aid, to proceed in search of relative bibliography or to make extensive researches in order to study analytically some cognitive object which concerns the case of the patient he examines.

The use of computational systems for the support of medical decision , it is possible to be proved very useful in practice. Systems of information technology can offer important help in various phases of the diagnostic process, particularly where the knowledge is possible to be structured and characterized by her objective character. The proposed system with the application of technologies of Artificial Intelligence and Data Bases will be developed in the mentioned before frames and it is expected to constitute an important aid in the diagnosis of children’s epilepsy ,particularly for the young persons and for specialized child neurologists. At the same time it will place the bases for the creation of corresponding inquiring infrastructure in the TEI of Crete, which in the future will improve under growth system with the use of additional technologies of Information technology and TeleInformatics and most probably will develop respectively new systems in other sectors of Medicine, where there is need of such tools and an extensive field of research.

For the system of support of diagnosis of children’s epilepsy is interested “Benizeleio Hospital”, Heraklion, Crete. This appears from his publications Sirs.s. Basilaki and from the presence of Mr P. Boytoyfianaki, child neurologist of Benizeleio Hospital, as collaborator of work.[:el]Υπάρχουν μεθοδολογίες ανάπτυξης και συντήρησης λογισμικού οι οποίες κατασκευάζουν προγράμματα στα οποία απεικονίζονται οι σχεδιαστικές αποφάσεις που έχουν χρησιμοποιηθεί για την κατασκευή τους. Τέτοια προγράμματα ή σχέδια προγραμμάτων μπορούν εύκολα να συντηρηθούν και να χρησιμοποιηθούν για την απόδειξη ιδιοτήτων τους όμως δεν είναι αποτελεσματικά. Σ’ αυτό το έργο θα μελετήσουμε και θα υλοποιήσουμε τεχνικές οι οποίες βελτιώνουν την αποτελεσματικότητα τέτοιων προγραμμάτων. Επιπλέον, θα μελετήσουμε το θέμα της επαλήθευσης λογισμικού και στοχεύουμε στην κατασκευή ενός επαληθευτή λογικών προγραμμάτων ο οποίος θα είναι ημιαυτόματος.

Επίσης, βασιζόμενοι σε τεχνικές Τεχνητής Νοημοσύνης, θα μελετήσουμε παράλληλες και κατανεμημένες δομές δεδομένων και αλγορίθμους για την πιστοποίηση συστημάτων και δικτύων σε chip. Οι βιβλιοθήκες που θ’ αναπτυχθούν θα επεκτείνουν υπάρχοντα εργαλεία πιστοποίησης όπως SCV (SystemC Verification Library) και Τestbuilder που βασίζονται στο πρότυπο μοντελοποίησης SystemC2.0.

Τέλος, σκοπεύουμε στην ανάπτυξη ενός συστήματος υποστήριξης αποφάσεων το οποίο έχει σαν στόχο να προσφέρει βοήθεια στην κατάταξη επιληπτικών κρίσεων που παρουσιάζονται στην παιδική ηλικία, σύμφωνα με την διεθνή ταξινόμηση για επιληπτικές κρίσεις. Παράλληλα θα αναπτυχθεί μία βάση δεδομένων στην οποία θα καταχωρούνται περιστατικά επιληψίας (φάκελος ασθενούς με όλα τα σχετικά δεδομένα). Η υλοποίηση των δύο προαναφερθέντων υποσυστημάτων και ο έλεγχος λειτουργίας αυτών σε πραγματικό περιβάλλον εργασίας, θα μας βοηθήσει να βγάλουμε συμπεράσματα σχετικά με την χρησιμότητα των αντίστοιχων συστημάτων σε διαγνωστικές διαδικασίες νευρολόγων για την ασθένεια της επιληψίας.[:]