Among Americans, heart disease is the leading cause of death. It is important to not only be aware of the many lives lost each year, but to also understand the economic impact. Each year, heart disease costs the U.S. healthcare system $214 billion dollars where 1 out of every 6 healthcare dollars is spent on cardiovascular disease. Using AI, Medical AI plans on combating these financial statistics while reducing costs and saving lives.
Medical AI, To develop AI algorithm using electrocardiogram (ECG) data, and makes it software to diagnosis and predict various diseases not only cardiac problem.
The medical team consists of BCD (Biosignal and Clinical Data), REA (Research for Evidence based AI-medical practice), and RFM (Research for Future Medicine) parts. Medical professionals from various fields who are dedicated to patient care in hospitals are playing an important role in creating medical-artificial intelligence convergence research and related software and services. We are thinking about how artificial intelligence can help health care providers care and treat patients. Our goal is to discover the “evidence” of “really needed” medical AI technology in the “medical field” and develop services to “help” healthcare providers and patients.
The SW team is composed of UX(User-experience)/Design, FE(Front-end), and BE(Back-end) parts. The UX/Design part studies service and screen planning, usability, and visualization, and the results are organically connected with FE and BE. The FE/BE parts review the demands connected to planning, design, and AI models and builds them in an optimal form. Team members alternate Evangelist roles over a period of time, exploring and disseminating new technologies within the team.
The AI team consists of Mlops, Dataops, and Research parts. The Mlops part develops a learning system to manage the entire lifecycle from model development to service, and the Dataops part designs pipelines by refining and validating ECG data collected through various channels. The Research part is researching/developing models with the best performance by utilizing these learning systems and large amounts of data.