AI for Disease Diagnosis

Employees of Vuno, a Korean medical device company, check Korea's first artificial intellgence-based medical device, the Vuno Med Bone Age.

As artificial intelligence (AI) has emerged as a vital part of the Fourth Industrial Revolution, many companies and researchers at home and abroad are taking on the development of AI-integrated services in the bio-health sector.

In particular, Korea can stand out in this sector as it has strong ICT infrastructure and reliable electronic medical records (EMRs) which are obligatory under a single universal healthcare system, analysts say.

Currently, nine products have been approved by the Korean Ministry of Food and Drug Safety as medical devices powered by AI. Since Vuno received a permit to sell the Vuno Med Bone Age in May last year, the first AI-based medical device in Korea, research on AI-based equipment has been on the rise. According to the ministry, approved AI-based clinical device trials also jumped from two last year to five this year.

It was IBM's Watson that ignited the development of medical devices using AI in the bio-health sector. The AI program helps doctors diagnose diseases by learning from cancer-related materials from more than 300 medical journals and more than 200 specialized medical books. In Korea, Gachon University’s Gil Hospital introduced the AI solution in August 2016. However, as Korean and U.S. patients generally suffer from different diseases, its effectiveness has been downplayed in Korea. Therefore, some Korean startups began to develop medical devices using AI for Korean patients.

In particular, the Vuno Med Bone Age, the first AI-based medical product in Korea, analyzes the wrist bones of a child in an X-ray photo and makes a diagnosis about the growth and development of the child. The AI program studies pictures of standardized carpal bones by machine learning and analyzes X-ray images of a child’s carpal bones. The Vuno Med Bone Age can predict future developmental conditions such as the height of the child, based on the gaps between bone nodes and the shapes and sizes of bones. Compared to the current method that doctors analyze by comparing 300 photographic images for 20 minutes, the Vuno Med Bone Age features close to 100 percent accuracy and analysis time of only about 20 seconds.

However, even after receiving approval from the Korean Ministry of Food and Drug Safety, the solution is faced with many obstacles before actual sales and distribution. The biggest stumbling block is medical fees. A diagnosis by the AI program and a diagnosis by a medical doctor bring the same charge to hospitals so many hospitals do not feel a strong need to introduce the Vuno Med Bone Age, medical industry experts say.

Copyright © BusinessKorea. Prohibited from unauthorized reproduction and redistribution