CareMate Clinical Assistant
CareMate is an innovative clinical assistant designed specifically for the Noongar Indigenous community in Australia. The system uses Natural Language Processing and Machine Learning to understand Noongar language input and provide clinical term extraction and prognosis prediction.
The Problem
Indigenous communities in Australia faced language barriers when accessing healthcare services. Medical professionals needed tools to understand symptoms described in Noongar language and provide appropriate clinical support.
The Solution
Developed a web-based chatbot using NLP techniques to extract clinical terms from Noongar language descriptions. Implemented a Random Forest classifier to predict prognosis based on symptom descriptions, bridging the language gap in healthcare delivery.
Key Features
- Noongar language to clinical term extraction
- Prognosis prediction using Random Forest classifier
- Interactive chatbot interface
- Symptom tracking and history
- Multi-language support (Noongar and English)
- Clinical decision support system
Challenges & Solutions
- Limited training data for Noongar language
- Mapping indigenous language terms to clinical terminology
- Ensuring cultural sensitivity in healthcare recommendations
- Balancing model accuracy with interpretability
Results & Impact
- Successfully demonstrated NLP capabilities for indigenous language
- Achieved 85% accuracy in clinical term extraction
- Positive feedback from community healthcare workers
- Potential for expansion to other indigenous languages
Project Info
Technologies
Frontend
Backend
Other
Tags
More Projects
Dingodesk
Cross-platform facility management solution for Australian small businesses. Features include onboarding, job scheduling, incident reporting, and job tracking.
Sysco Shop Global Platform
Food-service collaboration platform serving users for Sysco Corporation in the USA. Optimized BFF and GraphQL queries reducing load times by 40%.
Monidas IoT Dashboard
Real-time monitoring system for IoT devices with alarms and live data visualization in chemical industrial hubs in Switzerland. Engineered WebSocket connections for real-time streaming.