Register for free to
view the full innovation profile.
Digitisation of tissue introduces colour artefacts. Errors are different scanner to scanner. Scanner-agnostic Sierra standardises WSI images to real pathology colours for AI
About
The importance of colour management as a fine edge tool that makes WSI devices and AI both commercially competitive and next-level reliable is emerging at rate. The ground truth colours in histology are real data, from real patients and R&D pipelines that require diagnostics based upon reality for accurate and reliable healthcare. By utilising colour management techniques that have the ability to create standardised and ground truth coloured images irrespective of WSI scanner source, AI will be able to universally make life-changing decisions with complete colour certainty The market problem that FFEI's patented Sierra technology addresses is inherent to all WSI scanners and the markets they are deployed in. By using a simple slide containing real histology colours, developed in collaboration with world-leading digital pathologists and QA experts, and a software package that simply integrates into pre-existing and new workflows, Sierra can standardise all image colour data to the reality of the original glass slide. AI can be comparable to pathologist or cytologist microscope diagnosis, whilst providing quantitative data on which GLP auditing and medical QA reporting can be based.
Key Benefits
All WSI scanners introduce colour errors to different degrees, vendor-to-vendor or different models same vendor - even individual scanners of the same model can even have differences, exacerbated further by scanner age and intensity of usage. Sierra is scanner-agnostic and allows for any number of WSI scanners to be used, irrelevant of manufacturer, model or age caveats, and assures the user that their tissue colour data will be both standardised and true to the real tissue sample. This means that QA processes can be implemented, scanner output can be calibrated to remove colour artefacts and images can be validated ahead of AI and human-displayed analyses. This provides ROI in infrastructure of pre-existing WSI portfolios, which can be brought into standardised alignment with a simple integration and without the need for replacing expensive scanner platforms. With higher accuracy image data, analysis pipelines are more reliable and closer to ground truth, and therefore efficiency is increased when generating data to take new pharmaceuticals or algorithms to market. The quantitative metrics produced by Sierra provides a hitherto unseen and competitive level of colour accuracy on which GLP reporting and auditing for FDA approaches can be developed.
Applications
Calibrate scanner output and standardise tissue colour data across multi-source, multi-location datasets. Validate WSI colour performance and remove potential errors in data pipeline ahead of AI analyses. Generate quantitative data to establish data accuracy and support QA reporting to international standards. Deploy corrective ICC profiles to standardise images as part of qualitative analysis. Create an audit trail for GLP, FDA and medical QA accreditation.