Lunit helps physicians make accurate and objective diagnoses

through machine learning technology.

Data-Driven Medicine

“Medicine is a science of uncertainty and an art of probability.”

- William Osler -

Medical diagnosis is based on knowledge that results from generalization process of clinical evidences. Thanks to the birth of evidence-based -medicine (EBM) in 1992, the diagnostic decision making process had been more objective and scientific. In this context, we saw a big potential that EBM’s virtue could be further extended if we put better generalization technologies and clinical big data together. That’s why we believe that deep learning is the right technology for pushing the medical diagnosis to the next level - data-driven medicine.


Using Deep Learning

We believe that machine learning technologies can extend human ability beyond limitation of time and physical constraints.

We use deep learning to generalize the medical data. Generalization is the core objective of machine learning, but the existing technologies does not exceed capability of human, thus its value proposition is highly limited. We observed that deep learning had a lot of potential to get over this limitation. That’s why we focus on deep learning.


What We Do

Based on our 4-year experience of cutting-edge deep learning technology, we have a number of record-breaking algorithms for general object detection and medical image analysis. Starting from medical imaging, we will expand our coverage to various domains in medical records.


Digital Chest X-ray


Digital Mammography


Digital Pathology


Fundus Photography (TBD)

Let the Data Decide

Big data is at the core of deep learning. In order to aggregate as much data as possible, we are working with major general hospitals in Seoul, South Korea. The scale of our medical consortium is big and still counting: 10K beds, 12M outpatients per year.



We are a team of mission-driven experts in various domains including deep learning and medicine.


Our Backers


Feel free to email us to give us suggestions for fancy idea, or to just say hello!