Crypts detection in microscopic images using hierarchical structures

Authors:
Cristian Smochina;Vasile Manta;Walter Kropatsch
Affiliations:
Faculty of Automatic Control and Computer Engineering, "Gheorghe Asachi" Technical University of Iasi, Romania;Faculty of Automatic Control and Computer Engineering, "Gheorghe Asachi" Technical University of Iasi, Romania;Pattern Recognition and Image Processing Group, Institute of Computer Graphics and Algorithms, Vienna University of Technology, Austria
Venue:
Pattern Recognition Letters
Year:
2013

Citing 4
Cited 0

Linear Time Euclidean Distance Algorithms

IEEE Transactions on Pattern Analysis and Machine Intelligence
A topological sampling theorem for Robust boundary reconstruction and image segmentation

Discrete Applied Mathematics
Graph-based tools for microscopic cellular image segmentation

Pattern Recognition
Semantic segmentation of microscopic images using a morphological hierarchy

CAIP'11 Proceedings of the 14th international conference on Computer analysis of images and patterns - Volume Part I

Quantified Score

Hi-index	0.10

Visualization

Abstract

This paper presents an extended and improved version of an automatic technique which robustly identifies the epithelial nuclei (crypt) against interstitial nuclei in microscopic images taken from colon tissues. The detection of the crypt inner boundary is performed using the closing morphological hierarchy. The disadvantages of this approach related to the execution time and the used memory are highlighted and the morphological pyramid is used instead due to its computational efficiency, the reduced amount of used memory and the increased robustness. An analysis of the two approaches is performed considering the number of processed pixels, the used memory and the complexity. The outer border is determined by the epithelial nuclei overlapped by the maximal isoline of the inner boundary. The percentage of the mis-segmented nuclei against epithelial nuclei per crypt is used to evaluate the proposed methods. The limitations are described in order to highlight the situations in which the current approaches do not provide suitable results.