CardiologistAI

An AI-driven remote pre-screening ecosystem that utilizes rPPG technology to analyze facial blood volume micro-expressions and acoustic breathing intervals as a defense against cardiovascular risks.

Comment

Inspiration

Cardiovascular disease remains the leading cause of global mortality. In developing nations like Indonesia, this threat is amplified by a dire shortage of healthcare specialists. Driven by rapid population growth and compounding disease burdens, the country is projected to short more than 70,000 specialist doctors within the next decade. Furthermore, geographical fragmentation prevents millions of people living in deep rural environments from ever seeing a specialist. We built CardiologistAI to serve as a digital bridge. Because the entire platform operates fully online via a standard web application, patients can evaluate their cardiovascular risk factors instantly from home.

What it does

CardiologistAI is a web prototype designed as a remote cardiovascular pre-screening digital ecosystem.

  • For Patients: Features a streamlined online health portal where patients can initiate a 30-second face and voice scan directly using their device's built-in camera and microphone from the comfort of their home.
  • Instant Visual Reporting: Upon scan completion, the application compiles a comprehensive, interactive healthcare report which maps vital cardiovascular indicators such as live heart-rate graphs, audio synchronization scores, and simulated AFib or CHF risk analyses.
  • For Doctors: A secure, integrated clinical dashboard allows assigned cardiologists to safely track, monitor, and audit the longitudinal screening history of their remote patient pipeline in real-time.

How we built it

  • Frontend Core: React 19 paired with TypeScript for robust component type safety and state management.
  • Build Tooling: Powered by Vite 8 to maximize hot-module reloading and instantaneous local server compiling speeds.
  • Styling: Structured with the brand-new Tailwind CSS v4 utilizing an elegant, design-system-first utility setup for maximum interface responsiveness.
  • Backend & DB: Supabase handles our complete user authentication flow, secure relational mapping, and lightning-fast Postgres storage. Data isolation and privacy are guaranteed at the database layer using strict Row Level Security (RLS) backend policies.
  • Data Visualization: Recharts was integrated to map and translate complex streaming cardiovascular metrics into accessible, clean data visualizations.

Challenges we ran into

  • Tackling the entire development pipeline completely alone under an incredibly tight hackathon deadline.
  • Lacking experience because this was our first hackathon.

Accomplishments that we're proud of

  • Successfully wired a relational database architecture between independent doctor pipelines and remote patient records.
  • Built a highly responsive, polished UI container.

What we learned

This project taught us the importance of time-management and to build directly for human impact. During this project, we also gained the experience of stepping into the fast-paced ecosystem of a hackathon for the first time.

What's next for CardiologistAI

Given more time, we would like to expand our scope by implementing the following:

  • Native Multi-Language Support: Fully implement localized language engines (specifically introducing English and Bahasa Indonesia toggles) to ensure the platform is instantly accessible to remote communities across diverse cultural demographics.
  • Training with Real Biological Data: Transition beyond our current high-fidelity mathematical mock engine. We plan to train and embed native machine learning classification models (TensorFlow.js) using authentic, clinical cardiovascular data sets to evaluate micro-expressions and acoustic frequencies for real-world diagnostic screenings.

Built With

Share this project:

Updates