Implementation Challenges in AI for Pathology

Updated: about 11 hours ago
Location: Melbourne, VICTORIA

In order to achieve ubiquitous success in AI for Pathology around the world (in particular in histopathology), a number of key technical issues and challenges need to be addressed. These - and potential remedial approaches - will be explored in this research program and they include (as examples): 

  • variability in staining outcomes across different stains and different sites (even within a given Pathology company or service)
  • prevention and management of artefacts in digital pathology slides (eg - poor focus, external artefacts, air bubbles) 
  • management of changes in AI model types and performance over time 
  • challenges in managing data volumes (especially given the 3D format), and impacts related to digital slide and storage formats
  • the relationship of these issues to Quality Assurance programs in laboratory medicine 


Similar Positions