A healthy nation is a wealthy nation. At Flapmax our mission is to enrich the lives of everyone on the planet and make the world a better place. We strive to work with academic and healthcare institutions, industry, governments, and communities to make an impact in the health and well being of everyone.
This involves collaborating through open research to deploy accessible and sustainable computing platforms for EHR (Electronic Healthcare Record) and for genetic diseases studies (e.g., cancer), all while involving underrepresented researchers throughout the entire research and development pipeline.
As the demand for COVID-19 vaccine continues to rise, we partnered with G&H Pharmacy, a local pharmacy serving the City of Pinellas Park and Tampa Bay area of Florida, to develop an online platform for COVID-19 vaccination. For more than a decade, G&H has provided healthcare and pharmaceutical services to both low-income and middleclass residents, including providing free home delivery services to senior citizens in underserved communities. Together, we developed a HIPPA-compliant database system for the vaccination platform, accelerating access to vaccines by the local communities.
Equitable access to healthcare means leaving absolutely no one behind. To enable underserved communities in remote areas and suburbs with access to healthcare services, connectivity issues must be addressed beforehand. We have designed a pilot 5G/Edge computing platform and developed Coral Imaging to accelerate 2D/3D medical-imaging applications. The combined solution aims to allow radiologist to achieve increased accuracy and precision via privacy-preserving AI (PPAI). Coral Imaging was a winning a solution in the 2021 AWS Verizon 5G Computing challenge. Flapmax received cash awards and reinvested 80% of that fund into R&D, leading to the development of a new Federated Learning framework that will facilitate multi-institutional collaboration at scale amongst healthcare institutions.
Federated Learning trains models from various institutions in parallel and aggregates the models on a dedicated aggregation server without ever transferring data. This privacy-preserving AI technique is especially important as regulatory agencies closely regulate the privacy of patient data, e.g., GDPR in Europe; HIPAA in the US. We collaborated with the Federated Tumor Segmentation (FeTS) Challenge and Washington University School of Medicine in St. Louis to develop a new open-source framework for FL, achieving near state-of-the-art performance for the task of brain tumor image segmentation, while being flexible to support future FL algorithms.
A major issue facing Electronic Health Records (EHR) systems is not hardware lags, but complexities of mapping EHR data to a real-time stream in a very complex software ecosystem. To address these challenges, a unified, coordinated data and computing platform is required. We are collaborating with industry and government to design an accessible peta-scale data processing platform that will facilitate the development of accurate, inclusive, privacy-preserving AI models and data sets for next-generation EHR and research.