Related Experiment Videos
Hue-saturation-density (HSD) model for stain recognition in digital images from transmitted light microscopy
J A van Der Laak1, M M Pahlplatz, A G Hanselaar
1Department of Pathology, University Hospital Nijmegen, Nijmegen, The Netherlands. j.vanderlaak@pathol.azn.nl
Cytometry
|March 30, 2000
Summary
A new hue-saturation-density (HSD) model standardizes stain recognition in digital pathology images. This method improves quantitative analysis of stained tissues by overcoming limitations of the red-green-blue (RGB) and hue-saturation-intensity (HSI) models.
Area of Science:
- Digital Pathology
- Computational Imaging
- Biomedical Optics
Background:
- Transmitted light microscopy is crucial for examining stained tissues in pathology.
- Digital image analysis offers quantitative insights into tissue alterations.
- Standardized stain recognition is essential for reproducible quantification, independent of staining density variations.
Purpose of the Study:
- To evaluate color models for standardized stain recognition in digital pathology.
- To address limitations in existing color models for quantitative analysis of stained tissues.
Main Methods:
- Comparison of red-green-blue (RGB), hue-saturation-intensity (HSI), and a novel hue-saturation-density (HSD) color models.
- Utilized computer simulations and experimental data from immuno-doublestained tissue sections.
- Developed the HSD transform by applying it to optical density values derived from RGB data.
Main Results:
- The RGB model mixes chromatic and intensity information, hindering standardization.
- The HSI model struggles to separate distinct stains.
- The HSD model provides a standardized, two-dimensional data space for distinguishing all possible stain combinations.
Conclusions:
- The RGB model requires complex algorithms for standardization.
- The HSI model is unsuitable for stain recognition in transmitted light microscopy.
- The HSD model demonstrates superior performance for standardized stain recognition in digital pathology.