AI to Diagnose Achilles Tendon Injuries
Combining MRI with the possibilities of artificial intelligence will allow to improve the diagnosis of Achilles tendon injuries, which are among the most common orthopaedic injuries. Tendon MRI is shortened to a few minutes, and the description generation should be a matter of seconds.
This achievement was presented at the press conference `INNO THINKING` at the Institute of Physics PAS by Bartosz Borucki, head of R&D lab at the University of Warsaw. According to Borucki, imaging equipment has become cheaper and was now widely available, imaging test limits have ceased to apply and doctors are ordering more and more of them. However, the problem is describing the obtained images. There are not enough radiologists, so the waiting time for description is getting longer.
While atypical and complicated pathologies require and will continue to require assessment by an experienced radiologist, relieving specialists in assessing the results of simple, routine tests seems necessary.
Scientists from the University of Warsaw focused on the Achilles tendon. As the name suggests, the largest of human tendons is a weak point, not only for high-performance athletes.
According to the organizers` press release, in the US and Europe there are approx, 200 Achilles tendon ruptures per year per 1 million population. Each such case requires diagnostics; approximately 14,000 Achilles tendons are scanned in Poland every year.
`Diagnostics based on medical imaging opens new possibilities in the field of treatment and selection of optimal methods of post-traumatic or post-operative care`, argues Bartosz Borucki, quoted in the release. `We have created a solution for the assessment of the Achilles tendon, which introduces automation, enabling the creation of objective radiological assessments based on the use of artificial intelligence. This is the first such solution in the world. We are convinced that our project will set new directions for the development of imaging diagnostics in orthopaedics and sports medicine, and will improve the time and effectiveness of diagnoses`, he adds.
Thanks to the use of artificial intelligence (deep networks), Achilles tendon magnetic resonance imaging time is shortened from about half an hour to a few minutes, and the description is generated in a matter of seconds.
The Smarter Achilles MRI project is currently in the pre-implementation phase; if it obtains the appropriate certificate, it can be commercialised and marketed soon. The solution can be integrated with existing software for radiologists. Other orthopaedic injuries, for example torn knee ligaments, could also be assessed by artificial intelligence, assisting radiologists in their daily work (but not replacing them).
PAP - Science in Poland, Paweł Wernicki
pmw/ ekr/ kap/