Lunit said on the 1st that it will build a foundation model-based chest X-ray report generation model with SimonMed Imaging, the largest private outpatient radiology network in the United States.
SimonMed is a large imaging chain that operates about 175 medical institutions in 11 U.S. states. SimonMed has actively adopted AI-based medical technology, and this year the two companies have continued active collaboration, including the introduction of Lunit's 3D AI breast cancer diagnosis solution "Lunit Insight DBT."
The core of this partnership is Lunit's "foundation model service (FMS)." A foundation model is trained on large-scale medical data so that AI is equipped with basic medical knowledge. Through FMS, it is a cloud-based platform that can be fine-tuned to match the patient data and environment each hospital holds. It can provide consistent AI reading results while reflecting hospital-specific reading styles and patient characteristics.
This collaboration with SimonMed is the first case in which FMS is actually applied. Lunit plans to release additional models related to mammography (MMG) and digital breast tomosynthesis (DBT) next year and gradually expand its business portfolio related to foundation models.
SimonMed will introduce FMS into its imaging reading workflow to establish an efficient chest X-ray reading and reporting system at 175 medical institutions.
Seo Bum-seok, head of Lunit, said, "Because the methods of radiology reading and reporting vary by country and institution, there has been steady demand for AI suited to real clinical settings," and added, "The biggest advantage of using Lunit FMS is that each medical institution can build and operate a model that reflects its own data in a short period of time."
John Simon, head of SimonMed, said, "We have continually sought ways to improve the consistency and accuracy of medical imaging readings," and added, "We expect the introduction of FMS will significantly contribute not only to streamlining workflows but also to creating a better care environment."