Knowledge-based expert system applied in oral health towards tobacco use prevention and smoking cessation.

Hanganu, Stela Carmen ; Hanganu, Lucian Constantin

1. INTRODUCTION

Within the present context of tobacco-induced epidemics globalization, we considered a large research study in oral health area, approached in the research project theme, to be extremely necessary for at least two essential reasons (Anderson et al., 2002; Bentz, et al., 2007):

* to create a theoretical and scientific database, given the complexity of the problem;

* to offer the scientific base required for the individualization and implementation of the targeted programs for tobacco use prevention and smoking cessation in Romania.

The project adds important contributions in:

* the training and support of all peoples involved in tobacco control activities, including research, implementation and evaluation;

* the gradual elaboration of a national system for the epidemiological surveillance of tobacco consumption and of social, economic and health indicators related to it;

* the identification and testing of the methods of accelerated scientific results transfer into practice using dental medicine as a model;

* the evaluation of cognitive factors involved in the development and implementation of smoking control innovative methods and strategies in the dental practice;

* Romania participation as a partner to the global network of scientific research in smoking area in order to contribute to field knowledge development and improved access to information and to facilitate partners communication.

With these objectives in view, Romanian health policy included new principles, such as: the government responsibility of giving priority to health and to the concern for the own citizens' life, the recognition of non-smoking as a norm and of citizens' right to a smokeless environment, their protection from the negative effects of the passive exposure to smoke, the necessity to decrease yearly the prevalence of smoking in all the segments of the society, including the implementation of the Romanian national program Stop Smoking in 2007 by the Ministry of Health.

2. FUNCTIONAL STRUCTURE OF THE KNOWLEDGE-BASED EXPERT SYSTEM

The knowledge-based expert system SANFACTOR aims mainly positive influence patterns of behavior. Based on information received from the user, SANFACTOR makes an individual report for that particular user. It also aims to monitor progress of the user to obtain desired changes in behavior, the ultimate goal being the tobacco prevention and smoking cessation (Hanganu & Murariu, 2007; Hanganu, 2008). The structure of the knowledge-based expert system SANFACTOR is shown in figure 1. The main modules of this knowledge-based expert system are the interface dialogue, inference engine and knowledge base. The interface dialog allows a dialogue with users (patients) during the consultation sessions (digitized questionnaires and their answers), and access to facts and basic knowledge in adding or updating knowledge. It also allows expert users (health professionals) to bring the knowledge in a form recognized by the system and update the knowledge base of expert system.

The inference engine takes knowledge of the facts used to build reasoning. It seeks a number of major objectives, such as: choice of control strategy depending on the problem of the current development, solving plan development problem as necessary, switching from one control strategy to another, implementation of the measures provided in the solving plan. Although the inference engine consists of a set of procedures in the normal word sense, how to use knowledge is not required by the program, but depends on the available knowledge.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

Due to huge volume of information, the major benefit of this system is that it can store and process data more accurately than any human expert. SANFACTOR expert system should be seen as a tool to extend the medical skills and not as a substitute for it.

From the functional point of view, in general, the expert system SANFACTOR requires:

* a knowledge base - collection of facts and rules on the tobacco prevention and smoking cessation;

* an inference engine - which uses inference procedures for solving problems based on facts and rules in knowledge base.

The workflow of the knowledge-based expert system SANFACTOR is shown in figure 2.

The main stages of SANFACTOR operation are:

* the pacients must answer to an electronic questionnaire, with one or more existing answers; the questionnaire can be completed with the doctor help, or remotely using the Internet, and transmitting the standard form; all responses will be recorded for subsequent analysis; type of questions, type of answers and form of presentation will be chosen with experts on tobacco prevention and smoking cessation. the questionnaires will contain questions related to smoking habit (number of cigarettes per day etc.), risk factors for smoking-associated diseases (diabetes, heart disease etc.), actions trigger smoking (when it is drinking a coffee or alcohol etc.), psychological barriers that discourage early smoking (fear of fattening etc.), risk factors that affect the capability to quit smoking and early stage pre-smoking;

* the expert system SANFACTOR transforms data previously entered into a series of scores, representing a plurality of variables extracted from the trans-theoretical model of change;

* the variables extracted above are assessed to give a result appropriate attitude and behavior of the user;

* it is generated a report containing appropriate treatments (where applicable), forecasting (in the form of percentages), comparisons with similar cases in which the change of behavior took place successfully; this report is based on today medical and psychological literature information, with personalized messages and techniques of self-guidance for early smoking cessation, including nicotine replacement solutions; the results prove to be useful for those less motivated users who need real examples to mobilize them (Reiter et al., 2003).

[FIGURE 3 OMITTED]

3. CONCLUSIONS

The realization and implementation for the first time in Romania of a knowledge-based expert system in oral health towards tobacco use prevention and smoking cessation appeal to techniques of knowledge gain in the field of doctor-patient dialogue using specific communication means based on Natural Language Generation (NLG). The database will be structured as a personalized information package allowing the approach of specific applications designed for the implementation of new preventive and interventional methods of tobacco use control and smoking cessation, as a feedback to tobacco-induced epidemics globalization. Thus, will be created conditions for a better collaboration between various research-development entities in oral and general health area such as sanitary authorities, universities of medicine, Ministry of Health, institutes of public health, hospitals, medical R&D institutes, non-governmental organizations, in order to ensure the highest living standards--figure 3.

4. ACKNOWLEDGMENTS

This research was conducted within the framework of the PNCDI II Project 42123/2008 for which the authors acknowledge financing authority--The Ministry of Education, Research and Innovation of Romanian Government.

5. REFERENCES

Anderson, J. E; Jorenby, D. E.; Scott, W. J & Fiore, M. C. (2002). Treating Tobacco Use and Dependence--An Evidence-Based Clinical Practice Guideline for Tobacco Cessation, CHEST, Vol. 121, No. 3, pp. 932-941, ISSN 0012-3692

Bentz, C. J.; K. Bayley, K. B; Bonin, K. E.; Fleming, L.; Hollis, J. F.; Hunt J. S.; LeBlanc, B.; McAfee, T.; Payne, N. & Siemienczuk J. (2007). Provider feedback to improve 5A's tobacco cessation in primary care: A cluster randomized clinical trial, Nicotine & Tobacco Research, Vol. 9, Nr. 3, pp. 341-349, ISSN 1462-2203

Hanganu, S. C. & Murariu, A. (2007). Study of Romanian students behavioral and attitudes concerning smoking and tobacco cessation, premise for a healthy lyfestyle. Romanian Medical and Surgical Journal, Vol. III, Nr. 1, Suppliment 1, 2007, pp. 59-65, ISSN 0048-7848

Hanganu, S. C. (2008). National partnership in oral health area on tobacco use prevention and smoking cessation premises of the integration of Romanian research in the European operational space, Romanian research project nr. 42123/2008 within the framework of the PNCDI II

Reiter, E.; Robertson, R. & Osman, L. M. (2003). Lessons from a Failure: Generating Tailored Smoking Cessation Letters, Artificial Intelligence, 144, pp. 41-58, ISSN 0268-7526