Nordic Symposium on Digital Pathology
Richard Salmon, Guest Speaker
Richard Salmon will present on the topic of Color Management in Digital Pathology in the Open Innovation Session 2.
At a Glance
- Along with structural and spatial information, colour is a main component of all tissue pathology data
- Errors in histopathology colour interpretation are introduced by WSI systems, resulting in deviation from ground truth data
- There is a need for a method to correct this pathology of digitisation in digital pathology
- FFEI have developed a slide-based device that facilitates a scanner-specific ICC profile to ensure uniform and correct WSI images
- Advances in colour management has implications for diagnostic QA and QC, WSI scanner production, pharmacological GLP and medical AI
Colour Management has implications across the breadth of WSI fields; ensuring ground-truth data for pathologists; quality and validity of images produced by WSI vendors; re-calibration of machines after service visits and uniformity across systems used in GLP drug discovery. Of growing importance is the validity of data used in 3rd party software and AI – the future of diagnosis lies with AI, and whilst the effects of colour error are still emerging, a simple tenet of analysis is ‘bad data in = bad data out’. Colour management with FFEI’s Sierra technology provides a simple and effective method for ensuring the colour component of human and AI analysis is performed to ground-truth accuracy, preventing the problem before it can take effect. This application is ready for adoption and adaptation with all arenas of WSI use, reflected by deployment around a UK NHS network seeking to provide QA and QC for AI in digital pathology.
To register for the conference please visit the Nordic Symposium website.