Technical Name AI Osteoporosis Screening Assistant
Project Operator Biomdcare Corporation
Project Host 張漢威
Summary
Biomdcare’s Osteoporosis AI Screening Assistant is the first ever AI model to use hip X-ray image as the source to perform fast screening on Osteoporosis.
With patented in-house Low-Resource-Computing technology, Biomdcare is able to start the training process as well as achieve high accuracy with much less samples. The Osteoporosis AI Screening Assistant has achieved 90% of accuracy.
This Osteoporosis AI Screening Assistant can be equipped on mobile X-ray buses to help expand the screening area to communities and rural areas.
It is adopted by Taipei Municipal Wanfang Hospital, and has joined force with the hospital’s骨鬆肌少症照護中心, the new center that has opened in Feb 2020, to offer public a more comprehensive care on osteoporosis.
Scientific Breakthrough
According to the Taiwanese Osteoporosis Association’s publication referring to the 2019 International Society for Clinical Densitometry (ISCD), The WHO international reference standard for osteoporosis diagnosis is a T-score of -2.5 or less at the femoral neck. The determination of bone mineral density (BMD) is based on the dual-energy X-ray absorptiometry (DXA) as the gold standard, and both the PA spine and hip should be the skeletal sites to measure. When testing can’t be done on the hip and PA spine, forearm BMD should be measured.
The AI Osteoporosis Screening Assistant determines the femoral neck T-score is either greater than -2.5 or less than/equal to -2.5 in the X-ray image by training the hip X-ray image compared with DXA T-score, and the accuracy has reached 90% the above.
Industrial Applicability
The Osteoporosis AI Screening Assistant uses hip X-ray images as identification targets, and can be used with existing X-ray equipment in hospitals. There is no hardware purchase cost and eliminate the problem of uneven distribution of devices between urban and rural areas, and the hip area conforms to osteoporosis diagnostic medical guidelines. It will fill the osteoporosis screening market gaps.
Keyword Deep learning Low resource computing Medical X-ray image analysis assistant Osteoporosis Osteoporotic fracture Bone density AI for medical imaging Medical imaging X-ray image Disease
Notes
  • Contact
other people also saw