shrutic794/HeartReach

Developed an AI-driven, multilingual triage assistant (HeartReach) for early detection of female-specific cardiovascular symptoms in low-resource and offline settings. Integrated culturally-aware, low-literacy voice and text interfaces using WHO/AHA-based logic to deliver personalized care guidance via SMS, WhatsApp, and offline modes. # HeartReach: AI-Powered Cardiovascular Triage & Support for Women in Remote Areas ## Overview **HeartReach** is a multilingual, voice-enabled triage assistant designed to detect early signs of cardiovascular disease (CVD) in women, especially in **rural and underserved communities**. Built for **offline**, **low-literacy**, and **low-connectivity** environments, it combines **edge AI**, **context-aware logic**, and **empathetic health dialogue** to empower users with timely, personalized health insights. --- ## Problem Statement Cardiovascular disease is the leading cause of death among women globally, yet it remains **underdiagnosed** due to: - Gender-biased symptom recognition - Limited access to screening in rural areas - Lack of culturally-tailored health tools **HeartReach** addresses these gaps by thinking *female-first*, *offline-first*, and *community-first*. --- ## Key Features - **Voice-Driven Interface**: Users answer simple, spoken questions in their native language. - **Multilingual Support**: English, Hindi, and Spanish with future support for Tamil and Bengali. - **AI Risk Prediction**: Calculates cardiovascular risk using parameters like age, BMI, lifestyle, hypertension, anxiety, and family history. - **Offline Functionality**: Fully operational without internet — ideal for field deployment. - **Care Guidance**: Based on WHO/AHA guidelines, users receive tailored next-step suggestions (e.g., visit clinic, rest, monitor again). - **Risk Visualization**: Intuitive, color-coded progress bar with optional text-to-speech. --- ## Technologies Used - **Android Studio**, **Java**, **Jetpack Compose** - **ML Model**: Logistic regression / ensemble models trained offline - **Text-to-Speech (TTS)** API – for audio guidance - **Room DB / SharedPreferences** – offline storage - **Multilingual JSON templates** – for question flow and responses --- ## 🩺 Sample Risk Factors Considered - Age, Height, Weight - Family history of CVD - Hypertensi …