Application of neural networks to the segmentation of microscopy images

Authors:
Ling Guan
Affiliations:
Department of Electrical and Computer Engineering, Ryerson University, Toronto, Canada MSB 2K3
Venue:
Biocomputing
Year:
2004

Citing 7
Cited 0

Digital image processing (2nd ed.)

Digital image processing (2nd ed.)
Self-organizing maps

Self-organizing maps
Adaptive image processing: a computational intelligence perspective

Adaptive image processing: a computational intelligence perspective
Computer Vision

Computer Vision
Self-evolving neural networks for rule-based data processing

IEEE Transactions on Signal Processing
Weight assignment for adaptive image restoration by neural networks

IEEE Transactions on Neural Networks
A neural learning approach for adaptive image restoration using a fuzzy model-based network architecture

IEEE Transactions on Neural Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

An investigation of intelligent image processing algorithms to segment chromosomes in three-dimensional (3D) microscopy images taken by a confocal light microscope is presented. The use of this confocal light microscope allows biologists to observe live (or preserved) dividing cells in 3D. However, the top and bottom surfaces of these image features are indistinct, therefore requiring feature enhancement and segmentation of the chromosomes. In the proposed approach, a model-based neural network is first used to improve the quality of the images, and then the newly proposed self-organizing tree map (SOTM) is applied to perform segmentation. Segmentation algorithms are developed to work both on 2D dataset, based on a projection of the three-dimensional dataset, and on 3D dataset directly. The 3D approach to segmenting individual chromosome features preserves the 3D orientations in relation to the surrounding cell volume. The proposed algorithms perform very satisfactorily in the 3D case. Examples are provided to demonstrate the performance of the proposed algorithms.