SKT launches Korea's first AI-based veterinary X-ray image diagnosis assistance service ‘X Caliber’
SKT launches Korea's first AI-based veterinary X-ray image diagnosis assistance service ‘X Caliber’
  • Yeon Choul-woong
  • 승인 2022.09.26 10:51
  • 댓글 0
이 기사를 공유합니다

SKT launches Korea's first AI-based veterinary X-ray image diagnosis assistance service ‘X Caliber’. / Courtesy of SKT

SK Telecom (SKT) announced today that it has launched Korea's first AI-based veterinary X-ray image diagnosis assistance service named "X Caliber" after unveiling it at the 2022 Seoul Veterinary Conference held at September 25.

X Caliber is a web-based service that uses AI to analyze X-ray images of dogs taken by a veterinarian and uploaded to the AI platform ‘X Caliber Vet AI’, and delivers the analysis results back to the veterinarian within 30 seconds. 

Since X Caliber uses cloud to store and retrieve data, there is no need to install a separate server within the hospital. In addition, as a web-based service, X Caliber can be easily upgraded and managed. Veterinarians can examine the results of the AI-based image diagnosis on their mobile devices or PCs anywhere, anytime. 

The disease detection rate of X Caliber stands at 84-86%. It also showed 97% accuracy in measuring Vertebral Heart Scale (VHS). (84% for detection of 10 different abnormal patterns from chest X-ray images of dogs. 86% for detection of 7 different abnormalities from musculoskeletal X-ray images of dogs). 

SKT was able to realize the high performance X Caliber by fully leveraging its knowhow and expertise in areas ranging from dataset development and storing to AI modeling and service application based on its AI Full Stack.

AI Full Stack is an integrated package of AI technologies that connects technology elements of many different layers - such as infrastructure, hardware and software - that are necessary for realizing an AI service. It has been adopted as an international standard work item by ITU-T. 

To develop high-quality data sets, SKT cooperated with five national universities in Korea. Data augmentation technology was applied to increase the amount and diversity of data sets and enhance AI performance in clinical settings. 

For AI modeling, SKT employed automated data labeling and lightweight AI models. When developing an AI model, there needs to be a process where data is categorized, labeled, and processed into an optimal form for AI learning. 
Through active learning, it automated the data learning process and developed high-performance veterinary diagnostics AI models and high-quality AI data sets for learning. Also by realizing lightweight AI models, SKT has made it possible for users to get the results in a speedy manner. 

Moreover, the AI learning environment applied to X Caliber are developed on the cloud, making it convenient for the company to update service quality and distribute additional features. 

Although X Caliber is currently offering image diagnosis service for only dogs, the company plans to expand the service to cats next year. It also plans to promote overseas distribution of X Caliber in Europe and Asia, where no similar service exists, while constantly improving the disease detection rate of the service through strengthened deep learning. 

X Caliber is now offered as a subscription service for KRW 300,000 per month after a month of free trial. 

“With AI technologies being developed and applied to the medical sector at a rapid pace, the commercialization of SKT’s X Caliber is expected to serve as an important turning point for the advancement of veterinary medicine,” said Lee Young-won, professor at Chungnam National University College of Veterinary Medicine. 

“We were able to develop Korea’s first AI-based veterinary image diagnosis assistance solution by combining our expertise in AI and high-quality data from five national universities,” Ha Min-yong, Chief Development Officer of SKT. “We will continue to expand the scope of diagnostics to provide better pet care services and improve the welfare of animals.”


댓글삭제
삭제한 댓글은 다시 복구할 수 없습니다.
그래도 삭제하시겠습니까?
댓글 0
댓글쓰기
계정을 선택하시면 로그인·계정인증을 통해
댓글을 남기실 수 있습니다.

  • ABOUT
  • CONTACT US
  • SIGN UP MEMBERSHIP
  • RSS
  • 2-D 678, National Assembly-daero, 36-gil, Yeongdeungpo-gu, Seoul, Korea (Postal code: 07257)
  • URL: www.koreaittimes.com | Editorial Div: 82-2-578- 0434 / 82-10-2442-9446 | North America Dept: 070-7008-0005 | Email: info@koreaittimes.com
  • Publisher and Editor in Chief: Monica Younsoo Chung | Chief Editorial Writer: Hyoung Joong Kim | Editor: Yeon Jin Jung
  • Juvenile Protection Manager: Choul Woong Yeon
  • Masthead: Korea IT Times. Copyright(C) Korea IT Times, All rights reserved.
ND소프트