A fast morphological reconstruction algorithm for MRI brain image simplification

Authors:
Bofeng Zhang;Hui Zhu;Anping Song;Wu Zhang
Affiliations:
School of Computer Engineering and Science, Shanghai University, Shanghai, China;School of Computer Engineering and Science, Shanghai University, Shanghai, China;School of Computer Engineering and Science, Shanghai University, Shanghai, China;School of Computer Engineering and Science, Shanghai University, Shanghai, China
Venue:
HPCA'09 Proceedings of the Second international conference on High Performance Computing and Applications
Year:
2009

Citing 6
Cited 0

A contour processing method for fast binary neighbourhood operations

Pattern Recognition Letters
Fast computation of morphological operations with arbitrary structuring elements

Pattern Recognition Letters
A constant-time algorithm for erosions/dilations with applications to morphological texture feature computation

Real-Time Imaging
Efficient morphological reconstruction: a downhill filter

Pattern Recognition Letters
Image Analysis and Mathematical Morphology

Image Analysis and Mathematical Morphology
Flat zones filtering, connected operators, and filters by reconstruction

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

The MRI (Magnetic Resonance Imaging) brain image simplification has become one of important pre-processing steps for medical researches and clinical applications. Usually, the fast hybrid reconstruction algorithm, one of the common algorithms of image simplification, requires multiple iterations of reconstruction, so the time complexity is high. In 2004, Robinson gave a downhill filter to overcome this shortage, but it usually remains brighter regions because this algorithm targets MRCP (Magnetic Resonance Cholangiopancreatography) data. In response to the characteristics of MRI brain image simplification, this paper proposes a fast morphological reconstruction algorithm. Firstly, regional maxima concept is introduced to modify the initialization condition, then improved downhill filter algorithm is applied to reconstruct image. Experimental results show that the mean reduction of the execution time is about 73.68% compared to the fast hybrid reconstruction algorithm, and it can achieve good result when it is used to the contour extraction from the MRI brain image.