A fuzzy rule - based expert system for asthma severity identification in emergency department

- A fuzzy rule-based expert system for asthma severity identification in emergency department.
- A FUZZY RULE-BASED EXPERT SYSTEM FOR ASTHMA SEVERITY IDENTIFICATION IN EMERGENCY DEPARTMENT.
- However, in reality, long waiting time and uncertainty in the diagnosis has affected the quality of ED services.
- The fuzzy set theory, known as Fuzzy Rule-based Expert System for Asthma Severity (FRESAS) determination is embedded into the expert system (ES) to assess the severity of asthma among patients in ED.
- manages the fuzziness of the patient’s information data, and determines the subjective judgment of medical practitioners’ level on eight criteria assessed in severity determination.
- The system evaluation is performed by using datasets that were extracted from the ED clerking notes from one of the hospitals in Northern Peninsular Malaysia..
- The emergency department (ED) is a unit in the hospital that assesses and treats unscheduled patients arriving for immediate treatment.
- Apart from conducting emergency treatments, accurate diagnosis is also a critical aspect of excellent services in the ED.
- This paper focuses on the development of an expert system (ES) for asthma severity identification in the ED.
- The asthma cases were chosen as the medical domain to be investigated, instead of the many other cases at the ED as the boundaries of asthma studies are carefully identified..
- Asthma is a well-known disease that is described as a condition when both chronic airways are inflamed, while increasing airway hyper-responsiveness that leads to the narrowing of the airways, thickening of the airway wall and increase of mucus formation.
- Therefore, the purpose of this paper is to develop a fuzzy rule-based expert system that is capable of determining the severity of asthma in the ED.
- In the following sections, literature review and the methodology applied in the study are presented.
- An ED is one of the most important department in a hospital that does not only receive asthma cases to be treated, but accepts various other diseases and emergency cases.
- During asthma severity identification, one of the challenges in asthma management is that the asthma is classified based on the historical symptom patterns and evidence of obstruction in variable expository airflow (Levy et al., 2009).
- Modelling Uncertainty in the Expert System.
- In the case of uncertainty modelling approach, various mathematical approaches have been developed and introduced, such as the Fuzzy set theory (Zadeh, 1965), rough set theory (Pawlak, 1982) and Bayesian network (Spiegelhalter, Myles, Jones, &.
- Moreover, the necessity of the fuzzification approach in medical diagnosis ES is to overcome the problem of capturing subjective or ambiguous evaluations that are made in medical diagnosis (Ali &.
- From literature, the utilisation of fuzzy rule-based method as a decision support system for disease diagnosis and identification were explored, and a comprehensive review has been conducted by several studies which indicated the extended use of this methodology in the medical area (Mishra, 2014.
- The employment of fuzzy ES in the asthma case is common with several published papers, specifically in dealing with asthma cases that have started during pre-treatment of medical consultation till the predictions on the fatal causes of asthma.
- It is the first interactive computer-aided intelligent diagnostic system for the investigation of the type of asthma evaluation and its degree of severity.
- Most of the aforementioned studies in the fuzzy ES field were conducted in different countries, however, limited studies have been carried out in the local context, especially in Malaysia.
- There are many suitable methods to develop an ES, however for the acute asthma problems, the rule-based deduction method is considered, due to the following existing criteria of the case problem:.
- In order to assess the asthma severity, small context dependence is employed as one of the criteria that is applied into the rule-based method in ES..
- The objective of self-reading is to identify key areas and examine case studies on the acute asthma cases, and to obtain a basic understanding of the medical terms used in asthma cases..
- Subsequently, a semi-structured interview method was applied with multiple interview sessions with three academicians that specialised in mathematics and expert system field, one respiratory specialist, one clinical researcher and the Head of Emergency Department from one of the hospitals in Northern Malaysia.
- The first meeting was conducted to identify the common practices in handling acute asthma cases in local ED and determine the most significant assessment used in the classification of asthma severity..
- This process is conducted to ensure the originality and validity of the knowledge that are implemented into the expert system.
- After a lengthy discussion, a list of eight input parameters were decided to be used in the model development for Fuzzy Rule-based Expert System for Asthma Severity determination (FRESAS) based on the main GINA guidelines, whereby the output parameter diagnoses the severity of acute asthma as either moderate or severe acute asthma..
- The knowledge gathered is presented in the form of a mathematical formulation..
- FRESAS is developed based on two integrated techniques, which are fuzzy set theory and rule- were decided to be used in the model development for Fuzzy Rule-based Expert System for Asthma.
- The knowledge gathered is presented in the form of a mathematical formulation.
- In the fuzzy-based approach, Figure 1 shows the components of the proposed FRESAS method that consists of a fuzzy knowledge base, an inference engine that includes fuzzification, fuzzy inference rule, and defuzzification (working memory and user interface).
- Each of the FRESAS component development are elaborated in the next section..
- Through the inclusion of the PEF variable, in common medical practice, the normal PEF value was analysed from the current EU standard (EN 13826) PEF graph, and the resulting value obtained was compared to the measured peak flow meter.
- The fuzzy knowledge base of FRESAS is developed with the help of the medical experts’ past experiences as well as references to the current asthma guideline.
- The final rules were written in the format of <IF (antecedent) THEN (consequent)>.
- An example of the resulting rule statement is as follows:.
- When the previous condition is satisfied, then the rule is triggered and the action of the consequent is performed.
- During the fuzzification process, the real-valued on eight inputs of information are transformed into a membership grade of the fuzzy sets.
- Furthermore, due to the natural description of certain fuzzy variables such as oxygen saturation levels for which the values are read in a range, the trapezoidal graph was determined to be the most suitable representation for oxygen saturation levels in the fuzzy model.
- The identification of the memberships of each linguistic variable for each input variables were then transferred into the Fuzzy Logic Designer toolbox in Matlab.
- Graph of the membership function for conscious level..
- Graph of the membership function for talk..
- Graph of the Membership Function for Conscious Level..
- Graph of the Membership Function for Talk..
- Graph of the membership function for sit..
- Graph of the membership function for accessory muscle..
- Graph of the membership.
- Graph of the.
- Graph of the Membership Function for Sit.
- Graph of the Membership Function for Accessory Muscle..
- Graph of the Membership Function for PEF..
- Graph of the Membership Function for Pulse Rate (PR).
- Graph of the membership function for blood oxygen.
- Graph of the membership function for respiratory rate (RR).
- An example of the fuzzy membership expression for conscious level is represented in Equation 1..
- At this stage, the system identifies the linguistic variable selected in the user interface and matches it with the x-axis of the membership function graph of the fuzzy inference system.
- The value from the x-axis of the membership function graph is then mapped on the y-axis to obtain the fuzzification value, i.e.
- the confident level of the linguistic variable chosen.
- The results of the fuzzification outputs will then undergo the fuzzy logic operation process..
- Graph of the Membership Function for Blood Oxygen Saturation Levels (SpO 2.
- Graph of the Membership Function for Respiratory Rate (RR).
- An example of the fuzzy membership expression for conscious level is represented in Equation 1.1..
- In a part of the inference engine, there is a fuzzy operator which is responsible for combining the input of the fuzzy system.
- Hence, the fuzzy operator applied is based on the multivalued logic specifically focusing on min operation where it will present only one value that represents the antecedent part of the rule.
- The intersection of the fuzzy set min operation for FRESAS is presented in equation (2)..
- The result of the fuzzy operation determines the single output for antecedent, and continues to evaluate the consequent part of the overall rule statement..
- Min method is used for the implication of the latter part which decreases the fuzzy set output.
- The implication is also applied to each of the rules in FRESAS.
- Therefore, in the process to defuzzify the results of FRESAS, the aggregation process is assigned to combine each output of the fuzzy set rule into a single fuzzy set.
- The input for the aggregation process is the list of the decreased output functions that are returned from the implication process for each of the rules.
- With regards to the inference engine, the fuzzy operator is responsible for combining the inputs in the fuzzy system.
- Hence, the fuzzy operator applied in this approach is based on the multivalued logic that is specifically focused on min operation, where it will present only one value that would serve as the antecedent part of the rule.
- The intersection of the fuzzy set min operation for FRESAS is presented in Equation (1.9)..
- The implication is also applied to each of the rules in FRESAS..
- computes the centre of the area under the curve into a single number..
- Let A be the point representing the COG of the fuzzy set, and ab is the interval of the graph.
- Graph of the Membership Function for Asthma Severity..
- FRESAS applied a widely utilized defuzzification method known as centroid, which computes the centre of the area under the curve into a single number..
- Let 𝐴𝐴 be the point representing the COG of the fuzzy set, and 𝑎𝑎𝑎𝑎 is the interval of the graph.
- After consultation with the experts, the patients are required to fill up all the items in the interface to obtain a valid result.
- On the other hand, for the model validation, 30 real cases were collected within a month, and used as evaluation data to validate the application and functionality of the model.
- From the data collected, 57% of the subjects are female while 43% are male.
- The aim of this study study is to design an expert system for acute asthma severity determination according to GINA 2017 guideline in the emergency department.
- The developed FRESAS was evaluated with real data of acute asthma obtained from the emergency department’s medical records to measure the performance and effectiveness of the system.
- The results of the developed approach suggests that the built expert system of FRESAS is able to provide reliable and accurate diagnosis and hence, acceptable for real-time implementation..
- sensitivity and specificity of the expert system reported by Zarandi et al..
- This study did not investigate the effectiveness of the FRESAS used for doctors to improve patient outcomes.
- However, the low adherence of the GINA guidelines may be due to the changes made based on clinical experiences of the patient’s current condition.
- Moreover, the introduction of the FRESAS diagnostic system is highly recommended for real-time application due to its ability to determine the severity of acute asthma in ED that strictly adheres to the GINA guidelines..
- The application of an expert system in diagnosing diseases in the ED is essential due to the special characteristic of ED that deals with time-sensitive and broad spectrum of life threatening cases.
- Several limitations were identified and would require further study before the system can be fully utilized in the real world.
- Further experimentation of the system is still required to determine the validity of the system, and to obtain precise and fast outputs that ensure time management and system acceptance among medical practitioners..
- Clinical practice guidelines in the emergency department.
- 10.1016/j..
- Top 5 Benefits of Clinical Decision Support in the ED..
- In Proceedings of the 13th WSEAS International Conference on Applied Mathematics (Math’08), 8, pp.
- RI: an expert in the computer systems domain..
- In Proceedings of the 1st Annual National Conference on Artificial Intelligence.
- Academic Emergency Medicine : Official Journal of the Society for Academic Emergency Medicine, 8(12), pp..
- Immunology : Official Publication of the American College of Allergy, Asthma, &.
- 10.1016/.
- fuzzy rule-based expert system for diagnosing asthma.
- H., Moin, M., &

Xem thử không khả dụng, vui lòng xem tại trang nguồn
hoặc xem Tóm tắt