Medevice--使用神经模糊专家系统的基于移动的常见人类疾病诊断(IJITCS-V11-N8-4

I.J.InformationTechnologyandComputerScience,2019,8,27-32PublishedOnlineAugust2019inMECS()DOI:10.5815/ijitcs.2019.08.04Copyright©2019MECSI.J.InformationTechnologyandComputerScience,2019,8,27-32MeDevice:AMobile–BasedDiagnosisofCommonHumanIllnessesusingNeuro–FuzzyExpertSystemJohairaU.LidasanNotreDameUniversity,CotabatoCity,9600,PhilippinesE-mail:[e-mail]ty,CotabatoCity,9600,PhilippinesE-mail:[e-mail]:25April2019;Accepted:23May2019;Published:08August2019Abstract—Feverisasignthatthebodyistryingtofightinfection.Itisusuallyaccompaniedbyvarioussicknessesorsymptomsthatsignalanotherillnessordisease.Diagnosingitaheadoftimeisessentialbecauseithastodowithhumanlifeandtodeterminewhattodotogetwell.MeDeviceisamobile-basedapplicationthatrunsinAndroiddevicesthatallowstheusertoenterthelevelsofhis/hersymptomsanddiagnosesthediseaseeitherasinfluenza,dengue,chickenpox,malaria,typhoidfever,measles,HepatitisAandpneumoniatogetherwithitsdetailsanditsfirstaidtreatment.ItaimsatprovidinganefficientdecisionsupportplatformtoaidpeoplewithfeverindiagnosingtheirdiseaseandwhetherornottoseekmedicalattentionespeciallyindevelopingcountrieslikethePhilippines.ThisapplicationisengineeredwiththeknowledgebaseandtheinferencemethodoffuzzylogicandexpertsystemwiththehelpofGradientDescentoptimizationalgorithmandbackpropagationneuralnetworktoachievetheoptimumvalueoftheerrorrate.Thisisessentialtoprovidetheapplicationwithahighaccuracyratewhichshowsduringtheconductoftestingoftheapplication.IndexTerms—Fever,Neuro-fuzzyexpert,hybridsystem,Gradientdescent,optimizationalgorithm.I.INTRODUCTIONFeverisanincreaseinbodytemperature.Usually,thisistheresultwhenthebodyistryingtofightaninfection.Itisusuallyaccompaniedbyvarioussicknessesorsymptomsthatmostoftenthannot,signalsanotherillnessordiseases.Therearealotofdiseasesthatareassociatedwithorhavebeencalledfeverthataffectsalmostallhumanbeings[1].Therefore,diagnosingthediseaseassociatedwithfeverisessentialbecauseithastodowithhumanlifeandtodeterminewhattodotogetwell.Theonlyproblemispeopletendtoavoidseekingmedicalhelp.Basedontheconductedresearches[2,3],theyprefertostayathomeandwaitfortheirillnesstoimproveovertimeespeciallyindevelopingcountries.Thisisnotgoodsinceitcouldmaketheillnessworstand/orinfectotherpeople.Therefore,havinganapplicationthatwillhelpthepeopleindiagnosingtheirdiseaseisofgreatadvantage.ManystudieshavebeenconductedtodiagnosediseaseswiththehelpofArtificialIntelligencesuchasdiagnosingdiabetes,depression,typhoidfever,leukemiaetc.Artificialintelligenceisawayofmakingacomputerinasimilarmannertheintelligenthumansthink[4].Ithastwogoals:oneistocreateintelligentsystemsthatemulatethedecision–makingthecapabilityofhumanexpertsandusuallyusedinmedicalfieldsandindiagnosticssystems.ThisisdonewiththeuseoftheIF–THEN–ELSEstatementsorrules.Fuzzylogicandexpertsystemsorknowledge–basedsystemsaretwoofthebranchesofArtificialIntelligencethatusethesestatements.Expertsystemsareapplicationsthatsolveproblemswithasetofrulesoralsoknownastheknowledgebaseintheformofaseriesofquestions.Fromthese,theexpertsystemcandiagnoseandinterprettheinputsfromtheuser[4].Fuzzylogic,ontheotherhand,usesmembershipfunctionsindiagnosingandinterpretinginputsfromtheuser.ItallowstheusertohaveaninputbasedondegreesoftruthratherthanBooleanlogicinputsorcrispvalues[5].ThesecondgoalofArtificialIntelligenceisachievedthroughmachinelearning.Thisreferstotheautomateddetectionofmeaningfulpatternsindatalikeinpatternrecognition[6].Itisusuallydonebyartificialneuralnetworkswhereinitemulatesthestructureofthenervecellsbyhavingdifferentlayerswithdifferentneuronsthathavethecapabilitytolearn.Fuzzylogicitselforexpertsystemorneuralnetworkarecommonalgorithmsusedinthemedicalfield.Thereareexistingstudiesconductedthatusedoneoracombinationofthesesystems.Thecombinationofthesethreeresultedinahybridsystemthatdiagnosesdifferentdiseases.Buttheproblemwiththesestudiesisits28MeDevice:AMobile–BasedDiagnosisofCommonHumanIllnessesusingNeuro–FuzzyExpertSystemCopyright©2019MECSI.J.InformationTechnologyandComputerScience,2019,8,27-32availabilitybecauseitsmainusersarenotthepatientitselfbutthemedicalpractitionersormedicalpersonnel.Italsousesadesktopplatformwhichmakesitnotconvenientforthecommonpeoplewhodoesnothavecomputersortheinternet.Therefore,thisresearchisahybridintelligentsystemengineeredbyaneuralnetwork,fuzzylogicandexpertsystemtechniquesforthediagnosisofthediseasesassociatedwithfever.Itisamobile-basedapplicationthataimsatprovidinganefficientdecisionsupportplatformtoaidpeoplewithfeverindiagnosingtheirdiseaseandwhetherornottoseekmedicalattentionespeciallyindevelopingcountrieslikethePhilippines.ItdiagnosescommonhealthproblemsexperiencedbyFilipinosbasedonthePhilippineHealthAdvisoriessuchasinfluenza,dengue,chickenpox,malaria,typhoidfever,measles,hepatitisAandpneumonia[7].Thefollowingaretheobjectivesoftheapplicationsystem:1.Toestablishtheknowledgeoftheexpertinthesystemgiventhelevelsofsymptomsfromtheuser2.Todevelopthear