Doctor AI - AI Healthcare Chatbot

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 12 Issue: 04 | Apr 2025 www.irjet.net p-ISSN: 2395-0072

Doctor AI - AI Healthcare Chatbot

Ayushi Varma1 , Aamir Siddiqui2, Bhumika Nalawade3, Cyrus Parave4, Prof. Priyanka Bhilare5

1²³´ Department of Computer Engineering, Rajiv Gandhi Institute of Technology, Mumbai 5Assistant Professor, Department of Computer Engineering, Rajiv Gandhi Institute of Technology, Mumbai

ABSTRACT - Healthcare organizations currently experience elevated patient needs together with limited resources coupled with urgent requirements for precise medical knowledge. The healthcare system built on traditional face-to-face methods encounters difficulties deliveringtimelycarebecauseitcannotaddresscurrent patient needs effectively. This approach results in extended wait times and achieves poor management efficiency and medical diagnosis errors. Doctor AI represents an AI-based healthcare chatbot which the project develops to improve healthcare service interactionforpatients.

Doctor AI uses natural language processing (NLP) for patients to recognize their symptoms before getting medical advice from the system while it directs them toward qualified healthcare providers for advanced evaluation. The system links patients to evidence-based medicaladviceandsuitablehealthcarefacilitiesthrough its access to medical databases which uses symptom reports to determine required treatment protocols. The implementation of telemedicine functions enables patients to schedule immediate sessions which provides urgentaccess tohealthcareservices.

Through its design the Doctor AI chatbot intends to enhance healthcare service accessibility while maximizingresourceeffectivenesstocreatebetterpatient outcomes. The advanced technology solution presents a major stride toward using technology resources to improvehealthcaredeliveryduringthecurrentindustry challenges.

Key Words: AI-based healthcare chatbot, Natural Language Processing (NLP), Symptom recognition, Medical advice, Healthcare providers, Medical databases, Treatment protocols, Telemedicine, Immediate access to healthcare, Healthcare service accessibility, Resource effectiveness

1. INTRODUCTION

The healthcare industry faces multiple major obstacles which make patient care management inefficient. Rising patientvolume,limitedmedicalhumanresourcesandurgent medical information requirements create substantial organizational obstacles. Healthcare services delivered throughtraditionalpatientappointmentsrequiresignificant systemic delays in addition to extensive wait times. The

present healthcare environment proves that advanced solutions must be developed to create more accessible healthcaredelivery.

ThehealthcaredomaincanbetransfoSrmedusingthepower of Artificial Intelligence (AI), Machine Learning (ML) and NaturalLanguageProcessing(NLP)technologies.Thevast medicaldatasetanalysisabilitiesofAImatchwellwithNLP technology'sabilitytogivepatientsgenuinedigitalsystem interactions which create smooth user experiences. The projectDoctorAIasanAI-basedhealthcarechatbotwhich improvespatientengagementtotheuserswhileproviding timelymedicalassistanceseamlessly.

DoctorAIgivespatientsthecapabilitytochecksymptoms themselves and obtain quick medical advice before providing them with professional healthcare provider recommendations. The chatbot works to boost healthcare efficiencythroughoptimizedpatientmanagementtogether withreal-timeconsultationslikegenericmedicineorwhich doctortoconsulttoinordertoprovideimmediateaccurate medicalsupporttoindividuals.

2. LITERATURE REVIEW

The paper [1] discusses a text-based chatbot designed to assistuserswithhealthcarequeries.Thesystemimplements keywordextractionfromitspre-defineddatabasetoreturn responsesbeforeforwardingcomplexinquiriestoanexpert system.Theautomatedsystemseekstoprovideimmediate responseswhichhelpcutdownhealthcareexpenseswhile makinghealthcaremoreaccessibletousers.However,ithas severallimitations.Thestandardizeddatabaseprovidesonly programmedmedicalquestionswhichresultsinwrongand datedinformation.Thesystemworkswithalgorithmsthat preceded programming yet faces problems with unclear statementsandsophisticatedmedicalvocabulary.Thesystem operates without real-time adaptation because it needs manualupdatingratherthandynamiclearningcapabilities. Unanswerable queries require time because the system depends on an expert database which results in delayed responses for urgent matters. The problems with data storagesecurityandinformationprivacycreateworriesfor users sharing important health data since they encounter weak encryption alongside unsatisfactory handling of sensitive medical information. These challenges limit the chatbot’s effectiveness as an autonomous healthcare assistant.

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 12 Issue: 04 | Apr 2025 www.irjet.net p-ISSN: 2395-0072

The paper [2] explores an AI chatbot designed to handle COVID-19 inquiries using NLP. This system delivers assessmentsaboutsymptomsalongsidepreventionstrategies andvirusstrainreportstogetherwithvaccinationdetailsfor healthcarestaffinruralregions.Thesystemperformsonly symptom assessments through pre-established programs instead of diagnostic capabilities. The system depends on fixed data resources that prevents it from delivering contemporary medical information. The system does not have AI learning features which results in unvarying responses. Data security issues and slow response times occurbecausethissystemdependsonexternalAPIs.

The paper [3] discusses a chatbot that provides medical advice using Machine Learning and NLP. The system evaluatespatientsymptomsbeforeevaluatingtheircondition and recommending both Homeopathy and Ayurvedic treatments.Thesystemforwardscriticalmedicalsituations todoctorsbutrespondswithinformationthatcomesfrom retrieval algorithms. Predefined medical data makes the system perform diagnosis inaccurately since it lacks the capacity for real-time learning. The system must receive periodicmanualupdatingbecauseitlackstheabilitytoadjust tofreshmedicaldiscoveries.Thebotfacesdifficultieswhen receivingunclearinquiriesbecauseitrespondsthroughpredeterminedstatements.Patientprivacysuffersfromsecurity risks generated by the program's incomplete encryption methodsandundefinedregulatoryrequirements.

Thepaper[4]exploresWoebot,anAI-basedmentalhealth chatbot providing Cognitive Behavioral Therapy through mental health chatbot services. It aids in minimizing depressionsymptomsandanxietyaswellassubstanceabuse symptomsbyofferingbothmoodmonitoringandtherapeutic advice.LiteraturereviewsdemonstrateWoebot'scapability to maintain user interest and deliver confidential digital therapeutic services. Even though Woebot functions with programmedresponsesitlacksthecapabilitiestorecognize emotional states like human therapists do. This system cannot respond to ongoing mental health emergency situations because it does not have crisis intervention functions. Studying Woebot shows biases because the researchaimsonlyatselectgroupsofpeoplewhodifferfrom one another. The lack of clear encryption practices and insufficientregulatorymeasuresconcerningsensitivemental healthinformationresultsinethicalandprivacyissues.

3. HOW DOCTOR AI ADDRESSES THE LIMITATIONS

1. EnhancedNaturalLanguageUnderstanding(NLU)

Doctor AI surpasses standard retrieval-based protocols frompreviouschatbotinvestigationsthroughNLPmethods for finding precise medical words and symptom combinationsalongwithdiseaseindicationsinuserinput. The tool analyzes user sentiment through Sentiment Analysis to provide emotional interactions that mimic humanbeingswhilemostchatbotsfallshort.

2. More Accurate Recommendations with Machine Learning&DeepLearning

Previous health systems operated with set predefined rules and basic diagnostic capabilities. Doctor AI applies SupervisedLearningandDeepLearningmodelsmadeupof RNNs and LSTMs for enhancing symptom classification alongside intent recognition. The system moves toward patient-specific digital healthcare because it modifies its responsesaccordingtoreceiveduserfeedback.

3. ImprovedMedicalGuidance&DecisionSupport

Theformermedicalchatbotsfaceddiagnosticlimitations yet Doctor AI advanced its diagnostic prowess thanks to SemanticNetworksandOntologiesforenhancedknowledge representation capabilities. The system enables medical prescriptionrecommendationsandspecialistreferralswhich createsalinkbetweenfundamentalhealthchecktoolsand qualifiedmedicalconsultations.

4. Incorporation of Health Tools (BMI Calculator & PersonalizedSuggestions)

DoctorAIprovidesaBMIcalculatortouserswhichhelps them calculate their body mass index and delivers individualized health guidance according to BMI measurements.Thetemporaryinformationmanagementby DoctorAIforreal-timeinteractionsduringsessionsoccurs withoutmaintainingpersistentuserdatastorage.

5. Privacy-FirstApproachwithNoDataStorage

Patientsdonotneedtoworryaboutpatientdatastorage withDoctorAIsincethesystemdeletesallpatientrecords following each session. The system provides protected confidentiality by removing risks that stem from unauthorizeddatabreachanddatamisuseevents.

4. PROPOSED SYSTEM

The proposed system Doctor AI conducts patient data analysis through AI and NLP techniques to deliver precise responses. The core framework of the chatbot includes various essential components which improve healthcare accessibility together with operational efficiency. The NLP system receives user symptoms together with sentiment analysis data to conduct disease diagnosis by processing trained datasets. A key component of this module allows correct interpretation of patient questions together with medicallypreciseresponsegeneration.

The system incorporates supervised and unsupervised learningtechniquesthroughitsmachinelearningalgorithms thatenhancediagnosisaccuracy.Semanticnetworksaswell asontologiesusedinknowledgerepresentationhelpdoctors makebettermedicalchoicesthroughinferenceandreasoning capabilities.

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 12 Issue: 04 | Apr 2025 www.irjet.net p-ISSN: 2395-0072

DoctorAIoperatestoprovidefastmedicalconsultationsby decreasingtheneedforhumaninvolvement.TheAI-driven system provides real-time patient interactions with optimized resources which lead to enhanced healthcare results because it enhances accessibility of healthcare servicestouserneeds.ByleveragingMachineLearningand NLP Doctor AI enhances patient engagement, optimizes healthcare resources, and ensures timely medical intervention.

ID TestCaseTitle TestInput Result

TC_1 NaturalLanguage Understanding Whatarethe symptomsofacold? Pass

TC_2 Medical Information Retrieval

Tellmeabout diabetes Pass

TC_3 MachineLearningBasedSymptom Matching Ihaveaheadache anddizziness Pass

TC_4 BMICalculation Userinputsheight: 170cm,weight:75kg Pass

5.

METHODOLOGY

Development of the chatbot system used a planned sequential process which enabled seamless integration of naturallanguageprocessingtogetherwithmachinelearning models. Several essential steps made up the development process.Obtainingmedicaldatasetsaswellasuserqueries and healthcare responses from authorized sources represented the first step in data collection and preprocessing.Subsequenttrainingrequiredthatthedata passedthroughaseriesofcleaningandtokenizationphases. A natural language processing engine received training before it could classify user input while also extracting medicaltermsalongwithsymptomcategorization.Afterthe development of the diagnostic inference model came the integration of machine learning technology that provided both symptom identification and specialist recommendations. The system incorporated a medication suggestion feature which contained an incorporated databaseofover-the-countermedicationstogeneratesimple treatmentrecommendationswithfullcompliancetomedical standards. The web-based platform combines a secure authenticationfeaturewithaBMIcomputingtoolandchat capabilities to provide users with an efficient interface. Evaluation and testing of the system took place through repeatedtestingtostrengthenaccuracyanduserexperience andenhancesystemreliability.

The bot relies on natural language understandingfrom AI processingandmachinelearningfordiagnosticguidance.An NLP pipeline carries out the workflow by performing

tokenizationfollowedbystemmingthenintentrecognition andentityextractiontoanalyzeuserinput.Aclassification model trained with expert specialists selects appropriate medical expertsfromobtainedsymptoms.Thepredefined queriesbenefitfromrule-basedprocessingasthismethod deliversstructuredandpreciseresponsestousers.

Thesystemusesmultiplestructuredmedicaldatasources includingpublicmedicaldatabaseswithverifiedsymptomdisease data as well as healthcare domain corpora which containpretrainedmodelsformedicalNLPtasksanduser inputprovidesreal-timemodelfine-tuningcapabilities.The preprocessing phase included data clear-up steps for inconsistency removal as well as consistency-based formattingwhilelemmatizationandtokenizationnormalized text pieces for NLP processing and feature engineering producedessentialmedicalparametersthatboostedmodel predictivepower.

6. SYSTEM ARCHITECTURE

Figure1:SystemArchitectureofDoctorAI

FirstlythesystemarchitectureoftheHealthcareAIChatbot is in accordance to a structured approach using natural language processing and machinelearning models for the smoothUXbetweentheusersandthesystem.Thechatbotis designed with several interconnected part that are interconnectedwithdataprocessing,responsemaking,and medicalhelpsupply.

Theuserinterfaceconsistsofaninputmoduleandoutput module: it allows users to input queries and display the responsesfromthechatbot.Astheuserinputisprocessed throughanaturallanguageunderstanding(NLU)enginein thefirstplace,theNLUengineidentifiesuserrequirements andbreakdownsresponsemeaning.Thiswillensurethatthe systemunderstandstheuserintentandgetsthemeaningof the medical terms from the question. With the chatbot’s responsegenerationcomponent,thechatbotwillgenerate appropriaterepliesfromitspoolofresponsesdependingon theinputanalyzedbyformeaningandrelevance.

The dialog management module is incorporated into the systemto handle userconfusions,processesthe data,and validatesanyerrorsthatcouldoccurduringuserinteraction. Thismoduleplaysavitalpartinkeepingthechatbotsmooth communicationwithusers,insucha waythatresponse is contextual in nature and accurate. The main execution

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 12 Issue: 04 | Apr 2025 www.irjet.net p-ISSN: 2395-0072

processanddatafetchingoperationsareperformedbythe backend module. It fetches required medical information fromthedatabase,processuserinputandcallappropriate commandstogivenecessaryresponses.Itreliesonawellstructured and well-maintained database, the database containing things such as verified symptom disease relationship,recommendedoverthecountermedicationetc. Thiswouldmakesurethatthechatbotgivesoutcorrectand updatedmedicaladvice.

This chatbot’s processing workflow is structured, which helpsthechatbotanalyzeandrespondtouserqueriesinthe best possible way. The first step is tokenization and then stemming, intent recognition, and entity extraction. With these steps,thesystemisabletodissect user input tothe meaningful parts, the symptoms categorization, and the extractionofthekeywordsusedinthemedicalterms.Using expertmedicaldata,aclassificationmodelthendecideson the most suitable specialist recommendations from the symptoms it identifies. Moreover, the chatbot also offers rule-basedprocessingforpre-definedqueriesprovidedthat structuredandpreciseresponsesarerequired.

System’s key feature is that it uses multiple structured medicaldatasourcessuchaspublicmedicaldatabasesand pretrained models for medical NLP tasks. Real time user input is processed and model is fine-tuned based on user inputtoimprovetheresponseaccuracy.Thedatacleaning stepsandotherpreprocessingstepsremoveinconsistencies andstandardizeformattingcomeunderthepreprocessing phase.

The Body Mass Index (BMI) calculator included in the chatbot allows users to input their height and weight and providethemwiththeirBodyMassIndex.Accordingtothe BMI results, the users are classified into groups such as under-weight,weightnormal,weightoverweightorweight obese,andmakespersonalhealthguidance.

Thehealthcareinformationthemedicaldatabaseprovidesis usedasananchorforthechatbotsdecisionmakingprocess. Thisdatabasehasallthecommonsymptoms,thedetail of themedicalconditionsaboutwhichrecommendationcomes intermsofstandardtreatment.Keepingthechatbotupdated with regular updates ensures it gives accurate healthcare guidance as per present day medical guidelines. Further included in the database are explanations of medical conditions,treatmentprotocols,andgeneralhealthadviceto supportuserinformeddecisionmaking.

Thechatbotthroughitsrecommendationsystemprovides users with healthcare suggestions from the medically approvedsources.Basedonthepredicteddiagnosis,doctor recommendationsareprovidedandadviceisgivenouton over-the-countermedicationsforcommonhealthconcerns. Inaddition,thechatbotsharesgeneralhealthguidelinesto bearesourceformaintainingwellnessandpreventivecare. An integration of a machine learning based diagnostic

module with a medical database guarantees recommendationstobereliableandindividual.

An operational workflow for the system through a structuredsequenceprovidesabetterimprovementofuser interaction. Once a user has submitted his query, the NLP engine retrieves relevant symptoms and detects his intension. Diagnostic module processes extracted information and predicts the possible medical conditions andgivesrecommendationswithrespecttothespecialistsin caseofmedicalconditions.Thechatbotcalculatestheuser’s BMIifnecessaryandevaluatestheuser’shealthstatusfrom a weight and height basis. The chatbot verifies through necessary analysis and then generates a response with a diagnosis,arecommendedspecialist,anditisrelevantwhen anyassociatedoverthecountermedication.Thisstructured delivery makes the users get well documented and userfriendlyinformationthroughweb-basedinterface.

The Healthcare AI Chatbot bridges the gap between users and the value of professional medical care by integrating diagnostic insights,expertrecommendationsandoverthe countermedicationsuggestions.Preliminaryguidancethatis symptomadaptedisprovidedthroughthesystemenabling proactivemanagementofhealthcare.Asthechatbottools continue to be updated and tested incrementally, the accuracy and effectiveness of it will improve so as to enhance user interactions to a great extent and provide reliablemedicalassistanceeachtime.

7. RESULTS

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 12 Issue: 04 | Apr 2025 www.irjet.net p-ISSN: 2395-0072

Figure4:Chatbot'sresponsehandlinguserqueriesand providinghealthrecommendations

TheFigure2,Figure3,Figure4presentstheHealthcareAI Chatbotasitinteractswithmultipleinputsfromusers.The chatbottakesuser-submittedsymptomsthroughitsinitial interaction followed by diagnostic assessment to give probable medical assessment before advising healthcare consultation or precautionary practices. This healthcare platformhelpspatientschoosetheproperOTCmedication for coughs and nasal congestion from multiple selection options depending on their specific symptoms. The AI system proves its worth by offering suitable healthcare responsestorequeststhatdonotalignwithmedicalneeds. Users who need an essay-writing service encounter respectful declining behavior from the chatbot which providesredirectstogeneralphysicianconsultationsabout theirhealthissues.Thechatbotexhibitsoutstandingmedical ethics together with dependable guidance because it only provides medical information which remains within professionalparameters.

8. CONCLUSIONS

The healthcare support provided by Doctor AI features major improvements over previous chatbot-based healthcaresolutionsbecauseithandlesessentiallimitations. TheknowledgebaseofDoctorAIoperatesthroughnatural language processing techniques with machine learning capabilities to give precise medical guidance that follows patient context while enhancing user satisfaction through sentiment analysis and dynamic learning models. The system ensures precise and appropriate medical care for patientsthroughitsfeaturesforreal-timeconsultationsand symptomassessmentanditsabilitytoreferindividualsto specialistmedicaladvice.

Doctor AI ensures medical data privacy and security by adopting a policy which prohibits data storage from its systems.Thesystemfeaturesasetofadditionalcapabilities forBMIassessmentandhealthinstructionsandemergency readiness functions that create an overall effective digital healthcaresolution.

Doctor AI raises the standard of AI-minded medical care throughitsadaptablesystemthatbringsintelligentsolutions to health assistance while respecting patient privacy. The

futureofdigitalhealthcarewillbestrengthenedbyDoctorAI because it will receive three key improvements which combinemultilingualsupportandenhancedAIdiagnostics anddeeperhealthcarenetworkintegration.

ACKNOWLEDGEMENT

TheauthorsexpresstheirsinceregratitudetoProf.Priyanka Bhilareforherinvaluableguidanceandsupportthroughout thisresearchproject.

REFERENCES

[1] L.Athota,V.K.Shukla,N.PandeyandA.Rana,"Chatbot for Healthcare System Using Artificial Intelligence," 2020 8th International Conference on Reliability, Infocom Technologies and Optimization (Trends and FutureDirections)(ICRITO),Noida,India,2020,pp.619622,doi:10.1109/ICRITO48877.2020.9197833.

[2] S.Singh,M.Kaur,P.TanwarandS.Sharma,"Designand Development of Conversational Chatbot for Covid-19 using NLP: an AI application," 2022 6th International Conference on Computing Methodologies and Communication(ICCMC),Erode,India,2022,pp.16541658,doi:10.1109/ICCMC53470.2022.9753893.

[3] Ramalingam,Jegadeesan&Srinivas,Dava&Nagappan, Umapathi & Ganesan, Karthick & Venkateswaran, Natesan. (2023). Section A-Research paper Personal Healthcare Chatbot for Medical Suggestions Using Artificial Intelligence and Machine Learning Eur. 12. 6004-6012.10.31838/ecb/2023.12.s3.670.

[4] Dhanjal, G. 2025. Harnessing Artificial Intelligence for Global Health Advancement. Journal of Data Analysis and Information Processing, 13(1). Retrieved from https://woebothealth.com/