Tumour hypoxia is an important biological feature that is very close related to vasculature, and it has been proved to play a crucial role in the radiation response of solid tumours. In this paper we present a novel image analysis technique for simultaneous tumour hypoxia grading and blood vessel detection in dual-stained tissue sections, originated from the bladder region of patients treated by radiotherapy. The K-Nearest Neighbour classification scheme is employed initially in order to label the image colour pixels. Classification is based on a training set selected from manually drawn regions corresponding to the biological patterns being segmented. For tissue section images presenting a low quality staining, some further processing is required to reject misclassified pixels. A series of specific task-oriented routines have been developed (texture analysis, fuzzy c-means clustering and edge detection), in order to improve the final segmentation result. Validation experiments indicate that the algorithm can robustly detect these biological features, even in tissue sections with very inhomogeneous staining. This approach has also been combined with other image analysis procedures to objectively obtain quantitative measurements of potential clinical interest.
Characterization of the proliferative activity of a tumor has been the subject of research for many years. The majority of the studies presented so far in the field of cytology and histology relates to the analysis of information from a limited number of cells, which are often easily distinguishable from the background and as well as from each other. The present paper introduces an automated image analysis technique for classification of cancer cell nuclei stained with proliferative markers. The images under processing were characterized by a high degree of complexity, containing considerable histological noise. The first step of the method aims to identify nuclear features of proliferating cells only, contained in large-scale histological images, using Principal Components Analysis (PCA). The histogram of the component that demonstrates the best contrast is processed appropriately for generating a binary image. Some standard morphological operations are then applied to remove any irrelevant structures and detect touching and/or overlapping nuclei. Two separate methods, Skeleton by Influence Zone and heuristic processing, are presented for segmentation of clustered cells. The algorithm was tested on tissue section images encountered in routine clinical practice with very encouraging results, after comparing image analysis and human observer cell counting.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.