-
E-mail
hnrclh@163.com
-
Phone
15039091755
-
Address
Hanhai North Jin, Wenhua Road North Third Ring, Jinshui District, Zhengzhou City
Henan Rongcheng United Technology Co., Ltd
How to evaluate the color reproduction accuracy of a slide scanner?
Date: 2025-12-12Read: 0
In the era of digital pathology,slide scannerThe generated whole slide image (WSI) is commonly used for remote diagnosis, AI training, and scientific analysis, and its color reproduction accuracy directly affects the reliability of pathological interpretation. If the HE staining image is biased towards red, blue, or has distorted contrast, it may mislead the diagnostic conclusion.
The evaluation methods mainly include the following three categories:
Standard color card comparison method:
Use internationally recognized IT8.7/2 or ColorChecker Classic color cards to create HE simulation slices, and compare them with standard RGB values after scanning. Focus on observing the Δ E color difference in the areas of hematoxylin (blue/purple) and eosin (pink) (preferably<5, not acceptable>10).
Pathological quality control film verification method:
Prepare a batch of standardized HE quality control slides with known staining characteristics (such as fixed tissues, uniform staining batches), and label typical areas by a senior pathologist. After scanning, multiple physicians will blindly evaluate the consistency between the digital image and the microscope, and calculate the Kappa value (preferably>0.8).
Built in calibration function of the instrument:
A good scanner is equipped with white balance correction, light source stability monitoring, and ICC Profile automatic loading function. Automatic color calibration should be performed once a month, and the deviation before and after calibration should be recorded.
The influencing factors need to be synchronously controlled:
Scanning environment with stable lighting conditions;
The thickness of the cover glass meets the standard (0.17 mm);
The staining batches are consistent to avoid interference from human staining differences in the evaluation.
Only by ensuring true color reproduction can digital pathology truly achieve "what you see is what you get" and lay a visual foundation for precision medicine.