Polish journal of pathology : official journal of the Polish Society of Pathologists | 2019
A robust and automated cell counting method in quantification of digital breast cancer immunohistochemistry images.
Abstract
Quantitative analysis of\xa0immunohistochemically stained breast cancer specimens by cell counting is important for prognosis and treatment planning. This paper presents a\xa0robust, accurate, and novel method to label immunopositive and immunonegative cells automatically. During preprocessing, we developed an\xa0adaptive method to correct the\xa0colour aberration caused by imaging conditions. Next, a\xa0pixel-level segmentation was performed on preprocessed images using a\xa0support vector machine with a\xa0radial basis function kernel in HSV colour space. The\xa0segmentation result was processed by mathematical morphology operations to correct error-segmented regions and extract the\xa0marker for each cell. Validation studies showed that the\xa0automated cell-counting method had divergences varying from -5.05% to 3.99% compared with manual counting by a\xa0pathologist, indicating considerable agreement of\xa0the\xa0present automated cell counting method with manual counting. Thus, this method can free pathologists from laborious work and can potentially improve the\xa0accuracy and the\xa0reproducibility of\xa0diagnosis.