(Date: July 27, 2022)
Crescom Co., Ltd. (CEO Jae Joon Lee),
specialized in artificial intelligence medical image analysis solution
development company, announced on 27th that it succeeded in developing an
artificial intelligence ankylosing spondylitis diagnosis deep learning model
and was published in “Therapeutic Advances in Musculoskeletal Disease” of a
renowned international academic journal in the field of musculoskeletal health.
In this paper, Professor Bonsan Koo of Seoul Paik Hospital participated as the
first author, and Professor Taehwan Kim and Professor Seunghun Lee of Hanyang
University and Professor Changho Kang of Anam Hospital of Korea University and
others participated as co-authors.
Ankylosing spondylitis is an inflammatory
disease, one of chronic spinal joint disease in which all joints between the
vertebrae solidify into one, and if left unattended, it is a serious disease in
which the spine cannot move due to fusion, which may interfere with daily life.
In this study, we developed an AI solution
for grading mSASSS severity with detection of each corner of the cervical spine
and lumbar vertebrae, which is the severity grade of ankylosing spondylitis.
The average accuracy, sensitivity, and specificity of the vertebral edge mSASSS
score were 91.6%, 80.3%, and 94.2%, respectively. This artificial intelligence
solution is highly evaluated as it was developed for the first time in the
world to classify the severity of ankylosing spondylitis with high accuracy and
was first published in a renowned international journal.
Jae Joon Lee, CEO of Crescom, who developed
the deep learning model, said, “We hope that through the artificial
intelligence deep learning model developed this time, it is expected that it
will help specialists and contribute to more convenient and accurate early
diagnosis of ankylosing spondylitis and accurate evaluation of the progress and
treatment effect.”
Meanwhile, the international journal
“Therapeutic Advances in Musculoskeletal Disease” in which this paper was
published is a prestigious academic journal in the field of musculoskeletal
medicine, with a 5-Year Impact Factor of 4.774 as of 2022.