• Polish
  • English

Search

Kalendarium SI

June 2010 July 2010 August 2010
Su Mo Tu We Th Fr Sa
Week 26 1 2 3
Week 27 4 5 6 7 8 9 10
Week 28 11 12 13 14 15 16 17
Week 29 18 19 20 21 22 23 24
Week 30 25 26 27 28 29 30 31
Home arrow Aitech Sphinx arrow Salomon
Salomon PDF Print E-mail

Diagnosis of affective disorders
and
Bank of knowledge about healed depressive episodes
Author: This email address is being protected from spam bots, you need Javascript enabled to view it

 

Full version on CD-ROM can be ordered by e-mail.

INTRODUCTION

Computer advisory system "Salomon" with rapid access to the Internet data is aimed at providing the newest psychiatric knowledge to the first contact physicians, who are treating most patients with depressive disorders. Specialised psychiatrists may apply the system as useful tool, shortening the time of patient's waiting for effective medical assistance. System should never replace proper psychiatric examination and aggravate mutual relation of physician and patient. System can't replace physician in difficult art of taking therapeutic and diagnostic decision, will not replace his knowledge and experience. Computer advisory system "Salomon" was created as application of Polish expert system shell PC-Shell, by AITECH firm. Classification ICD-10, examination-diagnosis criteria for classification ICD-10, principles of classic binary logic (0,1 or "yes", "no") as well as of "fuzzy" (multi-value) logic, HDRS (Hamilton Depression Rating Scale), Pu¿yñski's mania rating scale and part of AMDP system are applied for defining rules in knowledge bases. Three databases were designed, containing data of the attending physicians, patients and methods of therapy, and data format for processing by neural network simulator was prepared, for aiding physician's decision, concerning method of treating depressive disorders, having character of mild, moderate or severe depressive episodes (single, recurrent or bipolar). System will be cost-free offered to all psychiatry clinics for critical review and obtaining information about subsequent consultations of patients with depressive disorders - necessary for creation of the future "bank of knowledge", concerning healed depression episodes.

GENERAL INFORMATION

System Salomon is the first Polish professional computer advisory system for diagnostics of affective disorders, based on scientific criteria and ICD-10 classification. The possibility of classification and evaluation of intensity of affective disorders, basing on the principles of fuzzy logic, apart of classification based on classic binary logic was introduced for the first time. Salomon system was created as application of the expert system shell PC-Shell for Windows, by AITECH firm (three times awarded on International Software Fair SOFTARG, used for commercial and scientific purposes). Program has modern hybrid architecture, suitable for newest advisory systems, and operates in Windows 3.1/Windows 95 environment. Salomon system was created as a result of author's previous experience, obtained in the course of performing his D.Sc. thesis, entitled "Attempts to improvement of diagnosis-therapy process of depressive syndromes by application of computer advisory system, based on different clinic classifications". The research project "Computer advisory system in diagnosis of depressions" awarded II prize in competition organised by International Association of Polish Psychiatrists.

APPLICATION

Typical fields of application of the system are as follows:

  • didactic of affective disorders ( medical education, pre- and postgraduate ),
  • assistance in first contact physician's and specialised psychiatrist's decision,
  • scientific research of affective disorders.

CLASSES OF PROBLEMS

  • "longitudinal" and "pointwise" classification of affective disorders,
  • evaluation of intensity of affective disorders in psychometric 5-stage scale,
  • analysis of introduced data (facts)
  • assistance in first contact physician's and specialised psychiatrist's decision

STRUCTURE

Salomon system is application of the PC-SHELL for Windows hybrid expert system shell. PC-SHELL is also blackboard system, what enables division of great knowledge base into smaller modules, topically oriented, named "knowledge sources". It is important and valuable property of the system.

REPRESENTATION OF KNOWLEDGE

Hybrid structure of system Salomon enables simultaneous application of different methods of knowledge representation:

  • declarative - in the form of facts and rules,
  • < object, attribute, value > triple,
  • imperative - in the form of algorithmic program,
  • knowledge in the textual form,
  • knowledge sources.

EXPLANATIONS

Explanations are very important, and even unique feature of expert systems. Broad range of explanation facilities, existing in modern advisory systems, is implemented in PC-SHELL system :

  • "how" explanations,
  • "why" explanations,
  • "what is" explanations,
  • "metaphors",
  • descriptions of facts.

"How" explanations are retrospective in their character and inform how given problem was solved by the system. "Why" explanations justify the reasons of questions asked by advisory system. "What is" explanations are textual information, concerning selected notions in the knowledge base. "Metaphors" are additional textual explanations, completing "how" explanations, introduced in PC-SHELL system. Metaphors are accessible after generation of diagnostic conclusion by the system and clicking push-button "how". Then diagnosis with preceding number of metaphor should be selected by mouse clicking, and only at that time the push-button "metaphor" should be clicked. Descriptions of facts show sources and methods of obtaining necessary information, and accessibility of "how" explanations.

DESCRIPTION OF SALOMON APPLICATION SYSTEM

Salomon system includes 5 knowledge bases:

  • mania.bw / mania /
  • mania_r.bw / mania flexible /
  • depresja.bw / depress /
  • in_z_af.bw / other affective disorders /
  • bank_dep.bw / bank /

Knowledge base mania.bw applies principles of classification ICD-10, and classic binary logic ( "Not"=0, "Yes"=1 ) for manic or hypomanic episodes with/without psychotic symptoms.

Knowledge base mania_r.bw applies principles of classification ICD-10, and fuzzy (multi-value) logic for manic or hypomanic episodes with/without psychotic symptoms. Patient's symptoms assume values from the set {0, 0.25, 0.5, 0.75}. These values are left limits of four ranges from < 0, 1 > ( or < 0 % , 100 % > ), where 0 (0%) - means the "normal state" ( not defined precisely ), 1 (100%) - extreme pathology (also not specified exactly), and medium values are defined by S. Pu¿yñski's mania rating scale.

Knowledge base depresja.bw applies principles of classification ICD-10, and classic binary logic ("Not"=0, "Yes"=1) for mild, moderate or severe depressive episodes with/without psychotic symptoms and with/without somatic syndrome.

Knowledge base in_z_af.bw (other affective disorders) includes classification principles for other, than above described, affective disorders : dysthymia, cyklothymia, bipolar affective disorder currently in remission or current episode mixed, other persistent mood disorders, mixed affective episode, recurrent brief depressive episode, and applies principles of classic binary logic ("Not"=0, "Yes"=1).

Knowledge base bank_dep.bw applies principles of classification ICD-10, and fuzzy (multi-value) logic for mild, moderate or severe depressive episodes with/without psychotic symptoms and with/without somatic syndrome. Patient's symptoms assume values from the set {0, 0.2, 0.4, 0.6, 0.8}. These values are left limits of five ranges from < 0, 1 > (or < 0 % , 100 % > ), where 0 (0%) - means the "normal state" ( not defined precisely ), 1 (100%) - extreme pathology ( also not specified exactly ), and medium values are defined by M. Hamilton's depression rating scale and system AMDP.

The data bases, containing data of the attending physicians, patients, methods of therapy, and subsequent consultations (examinations of patient's psychic state) are designed. They will be used for future bank of knowledge of healed depressive episodes.

The author suggests introduction of values characterising intensity of depressive symptoms in the course of classification of disorder, instead of its classification by the rules of classic binary logic ("Yes"=1, "Not"=0) and next application of different psychometric scales. As objective criterion of effective cure ( besides subjective, reported by the patient ), instead of reduction of number of points on the depression rating scale, author proposed the state, when patient's symptoms don't fulfil criteria of depressive episode of any intensity, although some of them may be still present.

Proper application of the system includes selection of registered physician's name from menu, introduction of patient's personal data (or his selection), and starting consultation. If introduced patient's symptoms fulfil diagnostic criteria of classification ICD-10 - the conclusion displayed by the computer will contain diagnosis with statistic number of illness and set of symptoms for so named neural network simulator, which in the future will automatically find similar healed patients, show the effective and ineffective methods of treatment, maximal dosing of applied medicines and the time of cure.

Succeeding consultations of the same patient should be performed in the case of taking decision, concerning change of method or place of therapy. In the case of ceasing of depressive episode, or when patient's symptoms haven't fulfilled criteria of depressive episode in the time of first examination - the only effect of consultation will be delivery of the set of symptoms for neural network simulator. In the case of healed depressive episode - subsequent consultations of the patient, with the use of program "Salomon" may be treated as valuable source of information, concerning method of treatment and dosing of medicines for preventing and supporting purposes.

INSTALLATION OF SALOMON SYSTEM - Internet version

Internet version is only in Polish version

When you get file from our server then unpack it and run instaluj.exe. After starting installation program - the dialogue window opens, enabling change of assumed directory of Salomon system. Installation begins after clicking push-button "Installation start". Clicking icon "Salomon", created after finishing installation , in group "Salomon 2.3" in "Program Manager", enables starting subsequent knowledge bases of Salomon system. Group "Salomon 2.3" contains also two auxiliary icons, enabling data bases archive.

BANK OF DEPRESSIVE DISORDERS

Application "Bank of depressive disorders" is aimed at gathering practical data, serving for creation future proper bank of knowledge about patients and methods of their treatment. This bank includes therefore the databases of patients and physicians, performing treatment. The effects of consultations with usage of advisory system for every patients, and therapy prescribed by physician are saved. Not less than four consultations included in the process of treatment of every patient are advised, if possible performed in regular time intervals.

Process of data gathering for this knowledge bank will last 6 months ( till end of June 1999 ). After this term the system Salomon will be inactive. Physicians and clinics, who want join the knowledge bank should send collected data to the author of the system (on diskette or by e-mail). These data will enable creation of common bank of knowledge. The persons participating in data gathering will have access to this bank.

Archiving of collected data is performed by the use of one of two archiving icons from group "Salomon 2.3" in "Program Manager", selected accordingly to possessed archiving program (arj/pkzip). The compressed file archiwum.arj or archiwum.zip should be sent on diskette, to :

Krzysztof Kielan
I Klinika Psychiatrii
Slaskiej Akademii Medycznej
ul. Grunwaldzka 48
42-700 Lubliniec
or by e-mail (as attached file ), to : This email address is being protected from spam bots, you need Javascript enabled to view it

OPERATION OF KNOWLEDGE BANK

The personal data of physician, using the bank of depressive disorders, should be written into database, when first starting the operation of program. These data have only informative character, enabling contact of the author of system with physician, for example when verification of reported results is needed.

Addition of new physician, using the bank, is performed by selecting option "New physician " from menu "Physician" and then introducing adequate data into opened window (presented on fig.1)

Introduced data are saved by selecting option "Save physician's data" in menu " Physician "

 


Fig. 1. Physician's database

 

New patient's data and results of his first examination are introduced in the window (see fig.2), opened by selecting option "New Patient" from menu "Patient". Only patient's 5-char identifiers are used in the system, instead of their names and addresses.

Introduced data are saved by selecting option "Save patient's data" in menu " Patient ".

 


Fig. 2. Patient's database

 

Real usage of the advisory system can be started after introduction of physician's and patient's data. Before beginning consultation the proper physician and patient should be selected from the list of registered physicians ( menu "Physician", option "Select physician" ) and patients ( menu "Patient", option "Select patient" ). Then new consultation is started by selecting option "Classification by fuzzy logic and ICD-10" from menu "Consultations".

Advisory system asks successive questions. The physician has to select proper answers from displayed list. Accessible explanations "What is" may be used for elucidating the asked questions. On the free screen space subsequent windows of firms sponsoring the project are opening. Exemplary consulting windows are shown on next figures.

 


Fig. 3. Exemplary consultation with advisory system

 



Fig. 4. Exemplary consultation with advisory system

 

Fig. 5. Exemplary consultation with advisory system

The conclusion - diagnosis generated by the system (one or more) will be displayed after answering all questions (fig.6).

 

Fig. 6. Conclusion of advisory system

"How" explanations, describing way of drawing conclusion, used by the system, are accessible in conclusion window. Exemplary "how" explanations, with auxiliary explanations of "metaphor" type, additionally describing the disorder, are presented on fig.7.

 


Fig. 7. "How" and "Metaphor" explanations

 

After closing the conclusion window - the system asks, if the performed consultation should be saved in the data base. Push-button "OK" should be clicked if the physician decides to save conclusion and prescribed therapy, push-button "Cancel" in the opposite case. The window for description of prescribed treatment opens after "OK" selection (fig.8). After finishing introduction of necessary data - the consultation should be saved by selecting option "Save consultation" from menu "Consultations".

 


Fig. 8. Therapy window

 

Bank of knowledge about depressive disorders in actual form enables auxiliary options listed below:

  • Reproduction of consultation (method, used by advisory system for attaining conclusion) - option "Reproduce concluding" in menu "Consultations".
  • Reproduction of history of patients therapy - option "Therapy history" in menu "Consultations".
  • View or change of physician's and patient's data - adequate options in menus "Physician" or "Patient"
  • Presentations of firms, sponsoring distribution of system Salomon, version 2.3, menu "Sponsors".

PRESENTATIONS OF SPONSORS

The author of the system wants to thank the following firms, helping in distribution of the system:

  • Eli Lilly, producer of Prozac,
  • Les Laboratoires Servier, producer of Coaxil,
  • Orion Pharma International, producer of Seronil.

Presentations of these firms are included in the system, in the form of Microsoft PowerPoint files, together with program PowerPoint Viewer, enabling displaying. The presentation may be viewed only in Windows 95 or newer environment, and computer advised for this purpose should be not worse than Pentium, 16 MB RAM.

The above listed firms are fully responsible for the contents and form of their presentations.

NECESSARY HARDWARE AND SOFTWARE:

System Salomon requires :
Computer: 486 or better.
Memory RAM: min. 8 MB.
Software: Windows 3.1 or Windows 95.

CONTACT WITH THE AUTHOR:

Krzysztof Kielan M.D.
I Psychiatry Clinic
Silesian Academy of Medicine
ul. Grunwaldzka 48
42-700 Lubliniec
tel./fax.: +48 34 3531768
e-mail : This email address is being protected from spam bots, you need Javascript enabled to view it

 

Zapisz się!

Zapisz sie na subskrypcję

Website awards

wyroz1.gif
MCportal.pl katalog stron

RSS News

Nominacje



Strona nominowana w Konkursie PERŁY INTERNETU - kataloog.info