CureHelp+ | AI-Powered Health Risk Analyzer

Your Personal Health Companion for Predictive Diagnostics and Medical Assistance

🌟 Overview

CureHelp+ is a Flask-based healthcare analytics platform that combines classical ML risk prediction, chest X-ray screening, medical report parsing, chatbot support, consultant discovery, a full-screen medical blog portal, and downloadable PDF reporting in one workflow.

It supports:

• Patient profile management with persistent storage and session sync
• Account authentication (OTP email verification + Google sign-in)
• Multi-disease risk prediction for tabular clinical inputs
• Pneumonia & Tuberculosis X-ray analysis via image upload endpoints
• Medical report autofill from uploaded CSV/PDF/XLS/XLSX files
• User dashboard overview with prediction/report/history aggregates
• Rule-based healthcare chatbot powered by curated datasets
• Consultant directory search (hospitals + doctors)
• Curated medical blog experience with interactive curved article navigator
• Admin dashboard for operational visibility and patient management

🔗 Project Links (Legacy)

• Live Link: https://www.curehelplus.me
• Azure Container Deployment Link: https://curehelplus.grayground-1f565d06.centralindia.azurecontainerapps.io
• Docker Image Link: https://hub.decker.com/r/asimhusain/curehelplus

✅ What’s New / Current Highlights

• Admin panel with login, dashboard metrics, and patient deletion actions
• Medical report upload with format validation and 200 MB size limit
• Enhanced PDF reports with risk gauges and detailed protocols
• Profile persistence layer with JSON and optional PostgreSQL backend
• OTP-based signup/login flows, password reset, and Google OAuth sign-in
• User profile photo upload and authenticated report/history tracking
• Dedicated /blog page with interactive 3D wheel article navigation UI
• Chest X-ray endpoints for pneumonia and tuberculosis risk scoring

🧠 Machine Learning Modules

Disease / Task	Input Type	Model/Approach	Output
Type-2 Diabetes	Tabular	Serialized sklearn pipeline	Probability (%)
Coronary Artery Disease	Tabular	Serialized sklearn pipeline	Probability (%)
Anemia	Tabular	Risk + type model stack	Probability (%) + severity/type
Pneumonia	X-ray Image	TensorFlow/Keras model (.keras)	Probability + Normal/Pneumonia
Tuberculosis	X-ray Image	PyTorch model (.pth)	Probability + Normal/Tuberculosis + confidence band

🧩 Core Features

1) Patient Profile Workflow

• Create profile using JSON or multipart form data
• Optional medical report upload during profile creation
• Session tracks active profile and disease predictions
• Search, list, and delete stored profiles

2) Prediction Workflow

• Tabular predictions:
  • Type-2 Diabetes
  • Coronary Artery Disease
  • Anemia
• Image predictions:
  • Pneumonia (multipart/form-data with image)
  • Tuberculosis (multipart/form-data with image)

3) Reports and Guidance

• Session prediction summary endpoint
• Filterable PDF export by selected diseases
• Risk-sensitive recommendation text generation

4) Chatbot and Consultant Directory

• Rule-based chatbot using bot_data/ datasets
• Cached responses (TTL controlled by environment variable)
• Consultant directory with hospitals/doctors and search support

5) Admin Panel

• Admin login/logout and protected routes
• Dashboard metrics: profile counts, prediction counts, high-risk indicators
• Disease and gender breakdowns
• Recent patient activity with delete operation

📁 Project Structure (High Level)

• app.py → Flask entry point, route definitions, model inference orchestration
• profile_manager.py → profile persistence manager (JSON/PostgreSQL + fallback)
• report_parser.py → extraction + field mapping for CSV/PDF/XLS/XLSX reports
• chatbot.py → dataset loading, query processing, response formatting, cache
• consultant.py → hospitals/doctors catalog and provider search
• makepdf.py → PDF generation with gauge visualization and recommendations
• admin/ → admin blueprint, auth flow, templates, dashboard logic
• models/ → trained model artifacts used at runtime
• bot_data/ → chatbot knowledge datasets
• templates/, static/ → frontend templates and static assets

🚀 Quick Start

Prerequisites

• Python 3.9 or higher
• pip
• Git
• Optional: Docker
• Optional: PostgreSQL (only if using DB-backed profile storage)

Installation

1. Clone the repository

git clone https://github.com/asimhusain-ai/CureHelpPlus.git
cd CureHelpPlus

2. Create and activate a virtual environment

python -m venv .venv
.venv\Scripts\activate      # Windows
source .venv/bin/activate    # macOS / Linux

3. Install dependencies

pip install --upgrade pip
pip install -r requirements.txt

4. Run the application

flask --app app run

5. Open in browser

• http://127.0.0.1:5000

⚙️ Environment Variables

Core App

Variable	Description	Default
CUREHELP_SECRET_KEY	Flask session secret key	curehelp-secret-key
TB_THRESHOLD	Tuberculosis decision threshold (0-1)	0.50
MODEL_WARMUP_ENABLED	Enable startup model warmup thread	true
MODEL_HEALTH_CHECK_INTERVAL_SECONDS	Periodic model health-check interval	300

Google OAuth

Variable	Description	Default
GOOGLE_CLIENT_ID	Google OAuth client ID	empty
GOOGLE_CLIENT_SECRET	Google OAuth client secret	empty
GOOGLE_REDIRECT_URI	Explicit callback URL override	empty
APP_BASE_URL	Base URL used to compose callback when redirect URI is not set	empty

Admin Authentication

Variable	Description	Default
CUREHELP_ADMIN_USER	Admin username	admin
CUREHELP_ADMIN_PASS	Admin password	curehelp

Chatbot

Variable	Description	Default
CHAT_CACHE_TTL_SECONDS	Chat response cache TTL in seconds	300

Profile Storage Backend

Variable	Description	Default
PROFILE_STORAGE_BACKEND	Force backend (json / postgres)	auto-detect
DATABASE_URL	PostgreSQL connection URL	empty
PROFILE_STORAGE_STRICT	Fail if postgres requested but unavailable	false
PROFILE_RUNTIME_JSON_FALLBACK	Enable runtime fallback to JSON on DB errors	true
PROFILE_AUTO_MIGRATE_JSON	Migrate JSON profiles into postgres on startup	true
PROFILE_DB_CONNECT_TIMEOUT	PostgreSQL connect timeout in seconds	2
PROFILE_DB_STATEMENT_TIMEOUT_MS	PostgreSQL statement timeout in milliseconds	3000
PROFILE_SYNC_QUEUE_MAXSIZE	Max queued async postgres sync operations	2000

These timeout controls help reduce login/reload delays when PostgreSQL is slow or temporarily unavailable by failing fast and allowing JSON fallback behavior.

Auth Email / OTP Delivery

Variable	Description	Default
SMTP_EMAIL_ADDRESS	Sender email/login for SMTP auth	empty
SMTP_APP_PASSWORD	SMTP app password / auth token	empty
SMTP_HOST	SMTP server host	smtp.gmail.com
SMTP_PORT	SMTP server port	465
SMTP_USE_SSL	Use implicit SSL SMTP connection	true
SMTP_USE_TLS	Upgrade plain SMTP with STARTTLS	false
SMTP_TIMEOUT_SECONDS	SMTP connect/send timeout	20

For Gmail, keep SMTP_USE_SSL=true and SMTP_PORT=465 with an app password. If OTP emails fail, auth APIs now return explicit SMTP errors.

🧾 Upload/Input Constraints

Medical Report Upload (POST /api/profile)

• Allowed formats: .csv, .pdf, .xls, .xlsx
• Maximum size: 200 MB

X-ray Upload (POST /api/pneumonia, POST /api/tuberculosis)

• Allowed formats: .jpg, .jpeg, .png
• Maximum size: 10 MB
• Request type: multipart/form-data

🔌 API Endpoints (Current)

General

• GET / → frontend landing page
• GET /blog → medical blog portal with interactive wheel navigator
• GET /api/config → baseline normal values for selected diseases
• GET /uploads/<path:subpath> → serve uploaded report/photo assets

Authentication

• GET /api/auth/status
• POST /api/auth/signup
• POST /api/auth/verify-otp
• POST /api/auth/login
• GET /verify-email
• POST /api/auth/resend-verification
• POST /api/auth/forgot-password
• POST /api/auth/reset-password
• POST /api/auth/logout
• POST /api/auth/logout-all
• GET /api/auth/profile
• PATCH /api/auth/profile
• GET /api/auth/google/start
• GET /api/auth/google/callback

Profile and Session

• POST /api/profile → create profile (+ optional report upload)
• GET /api/profile → get currently active profile from session
• GET /api/profiles?q=<name> → list/search profiles
• DELETE /api/profiles/<profile_id> → delete profile (if not active)
• POST /api/reset → clear session profile and predictions
• POST /api/profile/upload-photo → upload/update authenticated user profile photo
• POST /api/auth/reports → upload authenticated medical report
• DELETE /api/auth/reports/<report_id> → delete authenticated medical report
• GET /api/dashboard/overview → authenticated dashboard summary and charts payload
• GET /api/auth/history/export → export authenticated activity history

Predictions

• POST /api/diabetes
• POST /api/heart
• POST /api/anemia
• POST /api/pneumonia (image upload)
• POST /api/tuberculosis (image upload)

Reporting, Chat, and Directory

• GET /api/report → prediction summary for active session
• GET /api/report/pdf?disease=Type-2 Diabetes,Coronary Artery Disease
• POST /api/chat → chatbot response
• GET /api/consultants?q=<text> → provider search
• GET /api/metrics/latency → p50/p95/p99 latency summary + model health snapshot

Admin

• GET/POST /admin/login
• POST /admin/logout
• GET /admin/home
• GET /admin/
• POST /admin/patients/<profile_id>/delete
• POST /admin/users/<user_id>/activate
• POST /admin/users/<user_id>/deactivate
• POST /admin/users/<user_id>/force-reset
• POST /admin/users/<user_id>/delete

📦 Required Runtime Artifacts

Ensure these model files are available under models/:

• diabetes_model.pkl
• diabetes_scaler.pkl
• heart_model.pkl
• heart_scaler.pkl
• anemia_risk_model.pkl
• anemia_type_model.pkl
• feature_scaler.pkl
• label_encoder.pkl
• pneumonia_model.keras
• tb_model.pth

Chatbot dataset files expected in bot_data/:

• Disease precaution.csv
• DiseaseAndSymptoms.csv
• medquad.csv
• humanqa.csv
• Optional: Final_Augmented.csv

🧪 Testing

Run all tests:

pytest

🐳 Docker

Build and run locally:

docker build -t curehelplus:latest .
docker run -p 5000:5000 curehelplus:latest

☁️ Deployment (Azure Container Apps - Legacy Flow)

1. Build image

docker build -t curehelplus:latest .

2. Push to Azure Container Registry

az acr login --name cureacr
docker tag curehelplus:latest cureacr.azurecr.io/curehelplus:latest
docker push cureacr.azurecr.io/curehelplus:latest

3. Deploy

az containerapp up --name curehelplus --resource-group curehelplus --location central-india --image cureacr.azurecr.io/curehelplus:latest --target-port 5000 --ingress external --environment managedEnvironment-curehelplus-ade7

🌟 Contributing

Contributions are welcome.

• Fork the repository
• Create a feature branch
• Add/modify tests where relevant
• Open a pull request with a clear description

⚠️ Medical Disclaimer

CureHelp+ is intended for informational and educational support only. It does not provide a medical diagnosis and is not a substitute for professional healthcare advice.

Author

Made with ❤️ by Asim Husain — https://www.asimhusain.dev