AI‑Powered Smart Health Assistant

Project Domain / Category:

Web App

Abstract / Introduction

Finding clear, trustworthy guidance when you’re worried about your health shouldn’t feel overwhelming. This project proposes an AI-powered smart health assistant that listens to how people actually describe their symptoms by text or optional voice, and translates that into likely conditions with strong, evidence-based next steps. Our goal is to reach around 80% top-3 accuracy on a curated symptom condition dataset, while offering practical triage advice (self-care, see a GP, or seek urgent care) and recommending the right specialist for 90%+ of valid queries. To keep the experience transparent, each result comes with an easy-to-read explanation (via SHAP/LIME) for 95%+ of outputs. The system is designed to feel fast and dependable responding in about 2.5 seconds on typical text queries and maintaining 99% uptime during the demo period. And because accessibility matters, the full interface is available in English and Urdu, with accurate voice transcription (target 90%+ on clean audio).

Safety note: This assistant offers preliminary guidance only and is not a medical diagnosis. A clear disclaimer appears on every screen, and users are encouraged to consult qualified healthcare professionals for clinical decisions.

Functional Requirements:

1. User Roles

Guest User: Submit symptoms, view results, and read basic tips.

            Registered User (Patient profile): Guest features plus saved history, preferences/language, and PDF export.

            Admin/Clinician Reviewer (optional): Manage knowledge base, view anonymized analytics,

approve content updates.

2. Core Use Cases

UC‑01: Submit Symptoms (Text)

            Input: Free‑text description; optional metadata (age, sex, duration, severity 1–5), risk flags (pregnancy, chronic disease).

            Process: NLP preprocessing → symptom entity extraction → feature vector → ML/DL classifier.

            Output: Top‑k likely conditions (with confidence), triage level (self‑care / see GP / urgent care), recommended specialist.

UC‑02: Submit Symptoms (Voice) — optional

            Input: Microphone audio (8–15 s).

            Process: Speech‑to‑Text → same pipeline as UC‑01.

            Output: Same as UC‑01.

UC‑03: Personalized Guidance

            Input: User profile (age/sex/history) and current symptoms.

            Process: Rule‑augmented post‑processor (e.g., if fever + age < 5 → pediatrician).

Page 16 of 167

3.  Output: Next steps, home‑care tips, red‑flag warnings, specialist type, and when to seek care.

UC‑04: Explainability Panel

            Input: Model output and feature representation.

            Process: SHAP/LIME explainer on the final model.

            Output: Top contributing symptoms/phrases, feature‑importance plot, and a short

rationale sentence.

UC‑05: Multilingual UI

            Process: UI strings via i18n; Urdu/English toggle with RTL support.

            Output: Full UI translated; model output labels localized.

UC‑06: History & Export (Registered)

            Actions: Save queries and outputs; export last N results as PDF; delete history (GDPR‑like).

            Output: Downloadable PDF with disclaimer and timestamp.

UC‑07: Admin Knowledge Base (optional)

            Actions: Add/edit symptom synonyms, map conditions→specialist, edit self‑care templates, review logs.

            Output: Versioned content with rollback.

UC‑08: Feedback Loop

            Input: 1–5 helpfulness rating and free‑text feedback.

            Output: Stored feedback; optional model‑monitoring dashboard.

3. Data & Models

            Datasets: Public symptom–condition mappings (e.g., Kaggle/academic) plus curated synonym lists.

            Preprocessing: Tokenization; symptom entity extraction; spelling normalization; optional Urdu transliteration support.

            Model Options:

            Baseline: Logistic Regression / SVM / RandomForest on multi‑label bag‑of‑symptoms.

            NLP‑Enhanced: Clinical‑style embeddings (e.g., sentence‑BERT) feeding a classifier head.

            Hybrid: Rule engine for red flags combined with ML predictions.

            Explainability: SHAP Tree/Kernal explainers for tabular; LIME/SHAP for text.

         Inputs / Outputs & Validation Inputs:

            Text: 5–500 characters; block PII (names/phone/address) with regex before storing.

            Voice: 16 kHz mono WAV/WEBM; max 20 s; transcribe then discard raw audio (configurable).

Outputs:

Top‑3 conditions with confidence (0–1), triage level, specialist type, explainability plot, guidance bullets, and disclaimer.

Validation & Errors:

            Empty/short text → prompt for more details.

            Unsafe content (self‑harm) → show emergency resources message.

            Ambiguous symptoms → ask clarifying follow‑ups (fever? duration?).

         Workflows

            Symptom Flow: Input → Preprocess → Predict → Post‑process (rules) → Explainability → Render

→ Save (if logged in).

Page 17 of 167