Project developed in the context of Google Cloud Day Lisbon Hackathon: Leveraging AI for Portugal. Six teams have participated, and a different challenge was given to each of them. Ours was tasked with creating a smart triage agent using Google Cloud products.
Description · Features · Benefits & Impact · Next steps · Future features · Credits
RITA – Robô Inteligente de Triagem Automatizada [Smart Robot for Automated Triage] is a smart triage agent based on the Manchester Triage System that is able to take natural language input (text or voice) and classify the priority of patient treatment, based on the described symptoms.
Designed to streamline the triage process in healthcare settings, this project was developed during a hackathon to address the pressing issue of long wait times and overcrowding in emergency services. By combining AI with clinical expertise, RITA provides an efficient triage process, reducing the workload of healthcare professionals.
Watch recorded desktop demo.
Natural Language Input: Supports both text and speech for seamless interaction.Real-Time Triage: Helps patients triage themselves before arriving at the emergency department, reducing non-emergent visits.Integration with Hospital Systems: Can be connected with electronic health records (EHR) and updates the patient’s status in the hospital system.AI-Driven Triage: Leverages the Manchester Triage System to classify patients into different urgency levels, ensuring appropriate care.Cloud-Based: Uses Google infrastructure for scalability, security, and updates.
| Social |
|---|
• No Call Queue: Unlike traditional emergency services or hotlines, RITA eliminates the need for waiting in call queues. • Shorter On-Site Wait Times: By triaging before arrival, patients can be seen faster. • Relieves Healthcare Staff: Allows professionals to focus on urgent cases by reducing the non-emergent patient influx. |
| Technical |
|---|
• Managed Infrastructure: Google ensures the system’s security, scalability, and regular updates. • Interoperability: RITA can be integrated with other virtual and human agents, as well as hospital EHR systems. • Continuous Improvement: Machine learning continuously enhances the accuracy of the triage process. |
Development: Refine the triage dialogue flow in collaboration with medical professionals and ethicists.Data Collection: Improve access to quality data and validate user inputs.Integration: Seamlessly integrate patient medical history and status updates into hospital systems.Multilingual Support: Eliminates language barriers by offering multiple languages.
Monitor Physiological Factors: Track real-time health data.Observe Symptoms: Allow the system to use AI and smartphone camera for visible symptom identification.Service Suggestions: Recommend nearby services based on distance and wait time.
To know more about this project, check our pitch slideshow.
Jorge Costa @kelho0812
Jorge Cruz @mjorgecruz
Luís Carvalho @luis-ffe
Rui Pires @Rui-Pedro-Pires
Teresa Chow (me)
And a special thanks to Frederico Carvalho, our mentor throughout this challenge, for his continued support.