I've been working more and more on digital pathology in the past year, and I'm looking forward to an AMA workshop on AI coding policy in Chicago on December 8 (hybrid/virtual), which will continue the discussion of September's "Tab 94" at the AMA CPT meeting.
Pathologist David Braxton has a cool essay posted at Linked-In today, discussing the recent Digital Pathology Association meeting and giving us some highlights.
AstraZeneca QCS (TROP2)
He writes, "AstraZeneca's QCS System for TROP2 was a standout." This system for companion diagnostics is undergoing FDA review, and uses "normalized membrane ratio" as a biomarker. As Braxton writes, this is new and important because, "It's literally impossible to score an NMR by eye."
PathAI and AIM-PD-L1 for rearding PDL1
Braxton then points us to PathAI's new algorithm for PDL1 ("AIM-PD-L1") which stratifies survival [due to drug response] better than human pathologists. To achieve this, it reclassifies 30% of cases.
See Braxton's full essay at Linked-In here:
## For Extra Reading
See my blog today on PathAI's and Artera's comments to CMS on how to valuation AI or SaaS-SaMD in pathology.
