This paper provides a description of application of the regular decision tree at the pre-segmentation stage of the breast cancer fine needle biopsy microscope image analysis. The purpose the application is to improve results of segmentation, which is the next stage of the process. The proposed approach consists in the classification of image pixels based on their colour. The result is the image with pixels colour representing the probability of association to one of three classes: the nucleus, the intemucleus matter and the background. The enhancement is evaluated by applying the thresholding segmentation to the resultant image and comparing it with the results of the application of the similar process without the pre-segmentation stage.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.