RO  EN
IMI/Publicaţii/CSJM/Ediţii/CSJM v.20, n.2 (59), 2012/

Medical Image Registration by means of a Bio-Inspired Optimization Strategy

Authors: Hariton Costin, Silviu Bejinariu
Keywords: medical imaging, image registration, soft computing, evolutionary strategies, bacterial foraging algorithm, global optimization.

Abstract

Medical imaging mainly treats and processes missing, ambiguous, complementary, redundant and distorted data. Biomedical image registration is the process of geometric overlaying or alignment of two or more 2D/3D images of the same scene, taken at different time slots, from different angles, and/or by different acquisition systems. In medical practice, it is becoming increasingly important in diagnosis, treatment planning, functional studies, computer-guided therapies, and in biomedical research. Technically, image registration implies a complex optimization of different parameters, performed at local or/and global levels. Local optimization methods frequently fail because functions of the involved metrics with respect to transformation parameters are generally nonconvex and irregular. Therefore, global methods are often required, at least at the beginning of the procedure. In this paper, a new evolutionary and bio-inspired approach - bacterial foraging optimization -- is adapted for single-slice to 3-D PET and CT multimodal image registration. Preliminary results of optimizing the normalized mutual information similarity metric validated the efficacy of the proposed method by using a freely available medical image database.

Hariton Costin1,2, Silviu Bejinariu2
1Faculty of Medical Bioengineering,
"Grigore T. Popa" Univ. of Medicine and Pharmacy,
Iaşi, Romania;
Str. M. Kogalniceanu No. 9-13, 700454,
Iaşi, Romania

2Institute of Computer Science of Romanian Academy,
Iaşi Branch, B-dul Carol I No. 11, 700506,
Iaşi, Romania,
E-mail:

Fulltext

Adobe PDF document0.38 Mb