[Gettyimages Bank]
AI combined with big data is found to be extremely effective in image-based medical diagnosis as the robots can analyze and diagnose diseases using computerized tomography (CT) or magnetic resonance imaging (MRI) images much faster than human doctors. The clinical accuracy of medical AI is higher than 72 percent, according to a peer-reviewed study authored by researchers from Babylon Health and University College London in 2020.
In many countries, including South Korea, AI only analyzes about 70 percent of medical images before human doctors review the analysis and diagnose diseases. However, AI-based medical diagnosis systems are widely being adopted because the process is hundreds of times faster than the process only carried out by human doctors and researchers.
SK C&C said that the company's "Medical Insight Plus Cerebral Hemorrhage," an AI-based brain hemorrhage image analysis solution, will be listed on Microsoft's "Azure for Healthcare," the American tech giant's cloud-based healthcare-specialized platform. Hospitals that wish to use SK C&C's solution can simply download it from Microsoft's cloud platform.
"The latest technologies in the fields of digital and healthcare will integrate to create new innovations and contribute to the acceleration of South Korea's ESG sectors," Microsoft Korea's healthcare division head Yoon Dong-jun said in a statement on February 2.
SK C&C's Medical Insight Plus Cerebral Hemorrhage solution was developed through a collaboration with Ajou University Medical Center based in Suwon south of Seoul in 2019. The solution uses CT scan images of a human brain to detect cerebral hemorrhage. SK C&C has developed an algorithm that analyzes shapes by the size of pixels. The hospital has provided CT brain scan images of some 1,400 patients.
SK C&C had previously provided the company's AI-based solution through a cloud-based marketplace. In 2021, the solution maker partnered with Google to target the domestic and global markets for software as a service (SaaS).
reporter
Park Sae-jin
swatchsjp@ajunews.com
