Inspiration

Access to healthcare is still a significant challenge for low-income individuals, especially in underserved communities. We wanted to create something simple yet impactful that connects people in need with those who want to give back, without relying on cumbersome government systems or complicated policies. At its core, VCare is about empowering local communities to support their own, leveraging technology to make healthcare both accessible and secure.

What it does

VCare is a platform designed to bridge the gap between patients who can't afford care and those who want to help. By combining community-driven donations with healthcare accessibility tools, we provide a seamless way to improve health outcomes for everyone involved. Here's how it works:

Pay-It-Forward Donations: Benefactors can donate small amounts, similar to a deductible, either during their hospital visit or through our platform. These donations are pooled into "pay-it-forward" funds for hospitals in need.

Helping Patients Find Care: Patients who need financial assistance use the app to locate hospitals with available funds, making it easier for them to book appointments without the burden of upfront costs.

Remote Sentiment Monitoring for Mental Health: For patients with mental health challenges, such as schizophrenia, VCare integrates a sentiment analysis feature that allows users to share daily updates through text, voice, or video. The AI analyzes these updates to track emotional trends and sends insights to doctors, enabling remote care and reducing the need for frequent visits. This feature is particularly helpful for homeless individuals or those without reliable transportation.

SOS Alerts for Immediate Community Support: The app includes an SOS button that alerts nearby users and hospitals in the case of emergencies, ensuring that help arrives faster. This feature creates a real-time safety net by involving the community in critical moments.

Streamlined Identity Verification: At the hospital, patients provide their Social Security Number (or equivalent ID), which is securely verified through our platform to confirm eligibility for assistance.

How We Built It

VCare was built with a combination of AI, backend infrastructure, and a user-friendly frontend to ensure seamless functionality, security, and scalability. Here's a breakdown of how we approached the technical aspects:

Frontend

• Frameworks and Libraries: We used React for the frontend, ensuring a responsive, intuitive, and accessible user experience. Tailwind CSS helped us quickly style the platform and maintain a clean design.
• Geolocation Integration: We incorporated the Google Maps API to allow patients to locate nearby hospitals offering pay-it-forward funds or those with available appointments.
• SOS Alerts: The SOS feature uses geolocation services combined with real-time notifications, powered by WebSockets, to alert nearby app users and healthcare providers.

Backend

• Language and Framework: We built the backend using Node.js with Express.js to handle API requests, manage business logic, and support scalability for a large user base.
• Database: Patient and donation data are stored in MongoDB, a NoSQL database ideal for handling diverse data formats like patient records, hospital details, and donation tracking.
• AI Models:
  • Sentiment Analysis: Leveraged TensorFlow and Natural Language Toolkit (NLTK) to train and deploy an AI model for sentiment analysis. This model processes patient-submitted text, voice, or video inputs and generates reports for doctors.
  • Donation Optimization: We built a custom machine learning algorithm using scikit-learn, which predicts donation allocation needs by analyzing historical data like hospital capacity, patient demographics, and local demand trends.

AI-Powered Features

1. Sentiment Analysis: The model was trained on publicly available mental health datasets to detect emotions and flag signs of distress. This feature is implemented on the backend, where submitted data is processed securely, and insights are shared with the patient’s assigned doctor.
2. Donation Distribution: Using historical and regional data, our AI optimizes fund allocation to ensure that hospitals with the greatest needs receive sufficient support.

Secure Identity Verification

• We implemented blockchain technology to securely verify patient eligibility for financial assistance. This decentralized approach ensures that personal data remains private while offering transparent access logs for hospitals.

Cybersecurity

• Encryption: All sensitive patient and donor data are encrypted using AES-256 during storage and transfer.
• Authentication: User authentication is implemented using OAuth 2.0, ensuring secure access for both patients and benefactors.
• Real-Time Alerts: The SOS feature uses WebSockets with secure token authentication to prevent unauthorized access or manipulation.

APIs and Third-Party Services

• Stripe API: Used for secure payment processing to collect donations from benefactors.
• Twilio: Integrated for sending SMS alerts for appointment confirmations, donation updates, and SOS notifications.
• Google Cloud Platform (GCP): Deployed the backend using GCP’s App Engine for scalability and low latency. GCP also powers the sentiment analysis model via its AI infrastructure.
• Firebase: Utilized for real-time database updates and push notifications for the app.

Integration with Hospitals

• We built RESTful APIs that integrate with hospital management systems, allowing hospitals to:
  1. Access patient records securely.
  2. Verify eligibility for pay-it-forward funds.
  3. Update fund usage and availability in real-time.

Mobile App Compatibility

• While we focused on building the web platform, we used React Native to develop a basic prototype of a mobile app for patients. This app includes appointment booking, sentiment analysis updates, and the SOS alert feature.

Challenges We Solved Technically

• Scalability: To ensure the platform can handle an increasing number of users, we implemented a microservices architecture for modular scaling.
• Real-Time Data: Using WebSockets and Firebase, we ensured real-time updates for appointment statuses, fund availability, and SOS alerts.
• Model Optimization: Fine-tuned AI models for faster response times, ensuring that sentiment analysis outputs are delivered almost instantly.

By combining healthcare-focused functionality with robust technical infrastructure, VCare ensures a secure, efficient, and impactful experience for patients, benefactors, and healthcare providers alike.

What's next for VCare

1.Filter hospitals based on the type of aid they offer, since some hospitals may not have or resources for specific health issues.

2.Integrate the TensorFlow textsum library to summarize medical documents for the less-educated or non-native English speakers.

3.Carpool option if an incoming patient sees that a low-income patient in their area has booked an appointment around the same time or soon after.

4.Kiosks near subway/metro stop for booking appointments for those without access to computers.

5.General Location Donation (GLD): in the app select a zip code to donate to, then an algorithm would dole out the donation to the hospitals in that zip code that needs it.

Built With

• android-studio
• flask