Johnson has signed up Australian software engineering company Max Kelsen to use artificial intelligence to speed up the supply of surgical instruments to hospitals, helping to alleviate the backlog of surgeries that built up during the COVID-19 pandemic.
A single operation can take 10 trays of complex surgical instruments, says Warwick Shaw, customer solutions partner, Johnson & Johnson Medical.
“But we only have a fixed number of trays that we can deliver to surgeries, and if hospitals are working on the weekend or extending their operating hours to clear the waitlist, then we have to turn our trays around faster.”Using the camera on an iPad, SAVI automatically identifies instruments as they are returned, does the required paperwork, and tells the operator which instruments need to be manually inspected for damage.
Nicholas Therkelsen-Terry says machine learning makes fewer errors because it never gets distracted by social media.The co-founder and CEO of Max Kelsen, Nicholas Therkelsen-Terry, said using machine learning to identify surgical tools was an unusually difficult problem because the process required 99.99 per cent accuracy. This was much higher accuracy than ML models were normally capable of.Johnson’s trial had validated SAVI as having only a one in 10,000 error rate in the real world.