Poster
# 29
Poster |
|
6th Internet World Congress for Biomedical Sciences |
Dariane Deschamps(1), Anita Maria da Rocha Fernandes(2)
(1)Science Computer Course. Universidade do Vale do Itajaí - UNIVALI CTTMar - - Itajaí. Brazil
(2)CTTMar. Universidade do Vale do Itajaí - UNIVALI CTTMar - Science Computer Course - Florianópolis. Brazil
|
 |
|
[Odontology] |
|
[New Technology] |
[Health Informatics] |
Stomatology is a science that studies mouth and ist diseases. Periodontology is one of Stomatology´s branch and it is dedicated to the study, prevention and treatment of diseases assault the tissues that cover the teeth (1,2).
Epidemiology studies in many different countries showed that cavity and periodontal diseases with a large number of occurence in the mouth. So, it is necessary to help the dentist in order to heal these diseases.
Nowadays, the dentist bases on radiographies to conclude the diagnosis (3). By the use of expert systems it is possible to stablish the periodontal disease in less time and with more accuracy (4). The system will aid the dentist in the decision make, because the diagnosis work with a combination of many factors and variables.
The goal of this paper is to present the development of an expert system to aid the dentist in the diagnosis of periodontal disease, based on fuzzy logic. The methodology used here is based on the aspects bellow:
- bibliographycal research about computational aspects involved in the project;
- information acquisition about expert systems and fuzzy logic;
- study about tools that were used to develop the system: ToolBook and Fuzzy Logic;
- knowledge acquisition about periodontal disease.
The expert system development was very complex, so it was necessary to stablish some phases to implemented it:
- knowledge acquisition;
- knowledge representation;
- system implementation using FuzzyCLIPS;
- database development with pacient information using Access;
- user interface development.
Knowledge Acquisition
The knowledge acquisition corresponds to capture the knowledge of an expert to do this it was necessary many interviews with the expert in periodontal diseases. After the interviews it was possible to stablish that:
- to classify a bucal disease like a periodontal disease, it is necessary three symptons: bone loose; mobility; gengival bleeding;
- periodontal disease can be divided in two sub sets: "periodontite" or "gengivite". These two sets are sub divided in other categories;
- the principal variables involved are: age, sex, sistemic disease, implant, iatrogenia, stress, vitamin deficiency, evolution, bone loose, mobility, gengival bleeding, reduction in host resistence, tooth, ache, hormonal disturbance, medicine.
The first stage in the interview with the expert was about the disease identification: periodontite or gengivite.
The second stage was about the identification of periodontite kind, that was based on patient age. According to rhe expert, if the patient is young, there is only one treatment, but if it is a child and if the child has a sistemic disease, it´s possible to occur two different types of treatment.If the patient is an adult, it´s necessary a classification more detailed.
The expert said that if it is detected gengivite, there is a classification to each sympton.
Knowledge Representation
The phase related to knowledge representation shows the expert knowledge like production rules (5,6). It was stablished six sets of rules that FuzzyCLIPS inference (7) machine will study. An example is:
IF
bleeding = YES
bone loose = YES
mobility = YES
THEN
disease = PERIODONTITE
Expert System Development
The system implementation was done under FuzzyCLIPS. This tool was choosen because it was necessary to work with Fuzzy logic (7), and the expert stablished one fuzzy variable. The first step to the system implementation was the definition of global variables, and after, it was necessary to prepare the fuzzification and the defuzzification.
Considering that the variables input is done using ToolBook, it is necessary a function that read the input values that are stored in perio.txt archieve.
The system is composed by 54 rules. An example of these rules is:
(defrule R8
(PERIODONTITE "pre puberal")
(SISTEMIC DISEASE "no")
->
(assert(TREATMENT "chemistry + mechanic")
(printout resp "Pre puberal periodontite")
Database Development with Pacient Information Using Access
The system database was done using Access. With the database it´s possible to store information about the patients, they history and the treatments. The dentist can include, exclude and alter the information. Some reports can be printed with the infomation that the dentist wants. Figures 1, 2 and 3 ilustrates the database interface.
Figure 1. Database input menu.
Figure 2. Data input
Figure 3. History report
User Interface Development Using ToolBook
The user interface was developed by ToolBook. ToolBook is a software that help´s the implementation of multimedia systems. It is based on scripts and the programming language used is OpenScript (8). Figure 4 and 5 shows some modules of the system made using Toolbook.
Figure 4. System principal screen.
Figure 5. Expert System data inputs.
In the end of this jobit was possible to conclude that Artificial Intelligence, more specifically, Expert Systems, that shows a great utility to dentists in the decision make. It is important to remember that fuzzy logic is am interesting tool to model uncerteinty, but in this project it was poor explored.
It was verified that shell FuzzyCLIPS is a powerfull tool to develop fuzzy expert systems, but its interface is very poor, so it is necessary the use of another tool, like ToolBook.
|
|
|
|
[Odontology] |
|
|
[New Technology] |
[Health Informatics] |
