AI-Based Cancer Diagnostic Tool Supports Pathologists to Boost Accuracy Efficiency and Access with Joseph Mossel Ibex Medical TRANSCRIPT
Download PDFSep 30, 2024
Joseph Mossel, CEO and Co-Founder of Ibex Medical Analytics is developing AI-based cancer diagnostic technology to help pathologists work faster and more accurately with tissue samples. This technology is being widely adopted as an artificial pathologist that can perform tasks more efficiently and aid the pathologist in reducing error rates, which is ever more important with the shortage of pathologists and increasing demand. Partnering with pharma companies, this technology has the potential to develop algorithms that can help determine the effectiveness of specific cancer drugs, expanding the application of AI beyond improving diagnostic accuracy.
Joseph explains, "Pathology labs know how to receive a biopsy and a tissue sample and then convert it into a slide. Then, it is diagnosed by the pathologist. Our technology, not only does it help the pathologist to do their work faster, but it also is embedded within the lab workflow. You can think of the lab as the kind of a factory here, and we help all of that to be more efficient and faster. In some cases, we can actually cut down the time to diagnosis by even a few days."
"Many studies that we’ve run with pathologists found that using our AI is better than just the AI and better than pathologists not using the AI. And you’ll ask soon better by how much. That depends on the specific pathologists and the lab. But we’ve seen good labs with an error rate of around 2% to 3%. And by error, I mean really missing a cancer diagnosis, so diagnosing a cancerous case as benign. And then the absolute worst we’ve ever seen was a lab where 20% of the cases were misdiagnosed, which was a bit shocking. With a pathologist using our algorithm, the advantage of the two working together is that the algorithm and the pathologist tend to make different mistakes. Working together, we get to a level where there’s maybe a handful of mistakes in every 10,000 cases."
#IbexMedicalAnalytics #CancerDiagnostics #Cancer #DigitalPathology #Pathology #AI #AIinPathology #PatientOutcomes
