Medical Information Research Based On Fuzzy Rough Set And Fuzzy Integral

As we know, the treatment of a patient is a fuzz process which usually can not be described by certain words or numbers. And that is why we tend to ask for an old and experienced doctor to cure us. However,sine the abnormity degree of the patient is a series of expert data given by the doctor which means we can't keep the value absolutely precise,if we intend to set up a medical decision model,theory and technology which is good at dealing with uncertain information is needed to be the basis of the design.In this paper,medical information mainly refer to the characters of the patient's symptoms and expert data. And rough set as well as fuzzy set have their own advantages in fuzzy information.If we want to combine the practical problem with theory, the conditions of the patient and the experienced data came from history information as well as expert's knowledge should be treated as a system.Then how to analyse the system is the key of our medical model.Null-additive fuzzy measure is a mathmethic describtion which reflects the changes when two factors work on an object together. The decisive function is not simply equally to the sum of those factors on a certain problem separately.When two factors have positive function, the conjunct effect is larger than one add another while negative fuction makes it smaller than their summationMedical model in this paper combines the theory of rough set and fuzz integral.It takes advantage of the character of rough set which dose not need ancestor knowledge, and does the pretreatment to the data so as to get rid of the redundant properties. With that work done, the corresponding rules can be achieved.The next step is to do the fuzzy treatment to the rule we get above and making fuzzy operation with the rules in the form of fuzz set and the patient's fuzzy symptoms.The result of the calculation can be seen as the gist of judging the type of disease. To some disease that can be divided to sub-type further, the sub-type differs from each other in the form of the diversities among certain pathology parameter. Fuzz integral has its advantages in continuity information expecially in weighing the performance of each guide line.The classic Sugeno integral shows its main character in getting the max value from each minimum group while Wang integral is used to get the most effective value .They describe the condition of disease in their own way and as the requirement of the medical model, some changes should be made to Wang integral,which changes its purpose of getting the most effective value to achieving the least effective value.It indicates the least possibility that the patient suffers from the certain disease.Fuzz integral combines each factor in the relationship structure and does not need to keep those factors separated from each other which makes the judging result appears to be more scientific.