Sunday, September 30, 2018

Very Brief Blog: Recent Headlines in Digital Pathology

A CEO in the field explained to me, a few years ago, one of the differences between Digital Pathology and Digital Radiology:  Radiology information (in MRI, CT, PET) starts as digital information.  If we printed out "films" from these modalities, in the 1980s or 1990s, it was a second-hand workaround for an era when fast high resolution workstations were not common.  On the other hand, tissue histology starts with glass slides, not digits.

In late September, digital pathology made the front page of the New York Times, but not necessarily good news.  This was in stories about the $25M raised by Paige.AI as part of an exclusive-access deal to case files at Memorial Sloan Kettering and the back stories about how the deal was made, who got stock shares, etc.  Entry point here.

Some other late September news in Digital Pathology with a few links.

  • Startup Proscia pulls in an $8M new funding round for digital pathology software and AI-assisted innovations.  Here.
  • Publication that neural network software can correctly diagnose a high proportion of common dermatology diagnoses.  
    • Trade journal here, research article Olsen et al. here.
  • HealthCare Dive article on barriers versus momentum in digital pathology, here.
    • Note that this is a cover article for a forthcoming market sizing report by Signify Research, essay here and report access via here.  
    • The summary insights may still be of interest, though.
    • Radiant also offers a market sizing report here.
For a 2018 trade journal article and additional linked entry points on artificial intelligence in radiology, here and here.

For a few updates on cervical pathology AI in early 2019, see here and here.  For a couple blogs at FlagshipBio on AI in pathology, here and here.  Another update, April 2019 CAP TODAY re digital Her2 IHC guidelines, here.