• How it works

    Sierra Colour Calibration

TRUTH, INTENDED COLOUR & OUTPUT COLOUR

The FDA’s Technical Performance Assessment of Digital Pathology Whole Slide Imaging Devices document, which forms part of their whole slide imaging guidelines, states that, “The WSI system should be tested with a target slide. The target slide should contain a set of measurable and representative colour patches. Ideally the colour patches should have similar spectral characteristics to stained tissue.”

The interactions of stains with tissue create a large gamut of colours relatively unique to pathology. In order to achieve maximum accuracy, colour calibration technologies need to account for this colourimetric biochemistry.

FFEI’s patented Sierra technology utilises a specific biopolymer that precisely mimics the behaviour of stained tissue, by binding biological stains in processes akin to tissue staining. This unique feature allows the creation of a set of coloured patches that have spectral transmittance which is near identical to the gamut of stained tissue samples, and therefore goes beyond the basic colour combinations recognised in standard RGB values. By using different stains at different concentrations, it is possible to produce a metamerism-free range of the most commonly encountered colour intensities in histology and cytology.

MEASURING THE TRUTH

The spectral data of each patch for each calibration slide that is produced can be obtained using a high-precision spectrophotometer, calibrated to standardised transmittance materials of the National Physics Laboratory, to measure absorbance and intensity of colour.  This relates to the FDA’s requirement that, “The truth of the colour patches should be measured with proper apparatuses separately.”

This data is unique to each individual slide, eliminating variation in production and rigorously ensuring precision.  This is used to create spectral data specific for each slide that represents the true colours measured.

SIERRA IN PRACTICE

A THREE STEP PROCESS

The calibration slide is scanned by the WSI instrument to be calibrated. The output file, including the data for colour intensities, is recorded by the scanner, which is usually based on standard, non-biological RGB values.

A comparison is made between the colour intensities recorded by the scanner and the true colour (original spectral data) of the calibration slide. This then provides a numerical difference that indicates the accuracy of each colour scanned (‘output colour’) relative to its true measurement (‘intended colour’), therefore representing that WSI systems error across the colour gamut of tissue pathology, and thus the scanners colour fidelity to the ‘truth’.

With the ‘error’ for each colour established, the WSI scanner’s generic RGB colour profiling can be updated to bring it in line with the device-specific ICC profile calculated from the slide’s values.

Crucially, this is a ‘cross-checking’ process that does not interfere with the tissue image data itself and therefore maintains all the pathological data recorded by the scanner for the pathologist. Rather, this process adds in extra parallel data (metadata) that allows viewer software specifically designed to interpret calculated ICC profiles to represent the true colour data.