Citations & Acknowledgements
HealthTrack AI leverages and builds upon the important work of many researchers, datasets, open-source projects, and technology partners. We are grateful for their contributions to the scientific, medical, and developer communities.
Dataset Acknowledgements
MIMIC-IV Dataset
MIMIC-IV (Medical Information Mart for Intensive Care IV) is a large, freely-available database comprising deidentified health-related data from patients who were admitted to the critical care units of the Beth Israel Deaconess Medical Center (BIDMC) in Boston, Massachusetts. It contains data from 2008-2019.
HealthTrack AI utilizes a curated subset of approximately 10,000 deidentified patient records from the MIMIC-IV v2.2 `hosp` (hospital admissions) and `note` (deidentified clinical notes) modules. This data is instrumental in training embedding models for our similar case search feature, allowing clinicians to find relevant real-world case precedents. We do not currently utilize the ICU-specific (e.g., `chartevents`) or Emergency Department (`ed`) modules for this core feature.
For more information, please visit the Official MIMIC Website.
How to Cite MIMIC-IV (v2.2):
PhysioNet
PhysioNet offers free web access to large collections of recorded physiologic signals (PhysioBank) and related open-source software (PhysioToolkit). We acknowledge PhysioNet for its role in hosting and disseminating valuable datasets like MIMIC-IV.
How to Cite PhysioNet:
Core Technologies & Partners
MongoDB Atlas Vector Search
HealthTrack AI's rapid similar case search is powered by MongoDB Atlas Vector Search. This technology enables us to perform efficient similarity searches across the embeddings generated from the MIMIC-IV dataset, providing clinicians with quick access to comparable patient scenarios. We thank MongoDB for providing the robust and scalable platform that underpins this critical feature.
Learn more about MongoDB Atlas Vector Search.
Google Cloud Vertex AI
Our advanced AI models for tasks such as differential diagnosis and SOAP note enhancement are developed and deployed using Google Cloud Vertex AI. This platform provides the cutting-edge infrastructure and tools necessary for building and scaling sophisticated AI solutions in healthcare. We acknowledge Google Cloud for their powerful Vertex AI platform.
Discover Google Cloud Vertex AI.
Open Source & Design Acknowledgements
This application is built with the help of numerous open-source projects and design resources, including but not limited to:
- Next.js by Vercel
- Tailwind CSS
- Framer Motion
- Lucide Icons
- React, TypeScript, and many other foundational libraries.
Project Origins & Hackathons
HealthTrack AI was initially conceptualized and developed as part of our participation in the AI In Action Google Hackathon and the Hack the Vibe hackathon. These events provided a valuable platform for innovation and rapid prototyping, shaping the foundation of this project.