Fuzzy Expert System

1044 Words5 Pages

1.3 LITERATURE REVIEW In Malaysia, heart disease is the major factor of mortality that can cause a sudden death to patient, based on World Heart Organization coronary heart disease take a 98.9 life per 100,000 in Malaysia in 2012 (Ang & Chan, 2016). As we can see this is a big number of death as Malaysia is a developed country. Coronary heart disease is effect by a substance that accumulated inside the artery, this substance also call as atheromatous. Thus, it can cause thickening on the artery wall or narrowing the artery (HU, M. ,2016). There are numerous factors that can cause heart disease, therefore in our research we propose to study on Fuzzy Expert method that can be used to diagnosis heart disease based on several factors such …show more content…

The attribute is the most meaningful to factors of heart disease, from this they defined it into fuzzy expert system designing, membership functions, fuzzy rule base, fuzzification and defuzzification. Authors describe clearly the membership function for each attribute. After all of the attribute have been defined, the combination of attribute based on 44 antecedents the result are compared to experts’ lab and idea. Using this system it improve more rather than previous system and more logic also efficient. The advantage of this system is, the patients itself can use the system tools. There is experimental proven that showed this system is doing better than no-expert urologist almost 94% similar to the expert did (Adeli & Neshat, …show more content…

In this study, there are three stages that have been proposed by the expert system. All of this stages has been illustrate in a flowchart. In which, for the first step, is prep-processing units for DHD signal. In this stages, they obtain all of dopler heart sound from ultrasound system that can be got from department of cardiology. According to result, the AIS with k-NN fuzzy expert is efficient in interpret the disease. it is reported that 95.9% sensitivity rate and 96% specify rate which can be compare more higher or produced similar result with previous method such as ANN and WPNN. This method also seem good rather than fuzzy expert system in which they just defined the attribute in membership function and crisp value (Sengur,

Open Document