Structure-Preserving Color Normalization and Sparse Stain Separation for Histological Images

Abstract

Staining and scanning of tissue samples for microscopic examination is fraught with undesirable color variations arising from differences in raw materials and manufacturing techniques of stain vendors, staining protocols of labs, and color responses of digital scanners. When comparing tissue samples, color normalization and stain separation of the tissue images can be helpful for both pathologists and software. Techniques that are used for natural images fail to utilize structural properties of stained tissue samples and produce undesirable color distortions. The stain concentration cannot be negative. Tissue samples are stained with only a few stains and most tissue regions are characterized by at most one effective stain. We model these physical phenomena that define the tissue structure by first decomposing images in an unsupervised manner into stain density maps that are sparse and non-negative. For a given image, we combine its stain density maps with stain color basis of a pathologist-preferred target image, thus altering only its color while preserving its structure described by the maps. Stain density correlation with ground truth and preference by pathologists were higher for images normalized using our method when compared to other alternatives. We also propose a computationally faster extension of this technique for large whole-slide images that selects an appropriate patch sample instead of using the entire image to compute the stain color basis.

Effects of different preprocessing algorithms on the prognostic value of breast tumour microscopic images

“…However, the degree of prognostic enhancement did associate with the tonal range width as the best‐performing gamma correction by the exponent of 0.1 produced the narrowest, whereas the weakest‐performing contrast‐related normalisation algorithms produced the widest tonal range modification. This result was very surprising because colour variation between histopathology images has long been a recognised problem in image analysis (Li & Plataniotis, ), with image normalisation by stretching of pixel intensity range as the main strategy for the image enhancement in various applications (Li & Plataniotis, ; Bejnordi et al ., ; Vahadane et al ., ). Taken together, based on our data we confirm the known prognostic benefit of image tonal histogram normalisation and contribute the novel finding that nonnormalising tonal compression performs even better.…”

An alternative reference space for H&E color normalization

“…Sethi, et al [17] used standard deviation to describe intra-image variability and demonstrated that the Khan, et al [16] and Vahadane et al [11] algorithms exhibited variability similar in magnitude to the original image’s variability, concluding that much of the information in the original images was not lost. We noted a similar property in our results.…”

“…Recent work has used color normalization as a tool to transform H&E images to conform to a standardized color target in a manner that minimizes the loss of useful information in the image. Many of these approaches have used intensity thresholding [8, 9], histogram normalization [10], stain separation [11–13], and color deconvolution [14–16] to characterize or normalize H&E images, with varied levels of success. These algorithms shared a common strategy in that normalization can be approached from the standpoint of color analysis: that the superposition of stains that form the color basis of an image can be deconstructed into its constituents and recombined to conform to a target.…”

Spatial Statistics for Segmenting Histological Structures in H&E Stained Tissue Images

“…Specifically, we asked an expert to select a bag of dominantly stained hematoxylin and eosin pixels. In the future, this process can be automated by calculating stain vectors for hematolyxin and eosin over a population of H&E slides and using these vectors to derive representative pixels [15], [16]. …”