Systematic comparison of mutual information-based methodologies for the multimodal register of medical images


Authors

  • Gustavo Adolfo Ospina Torres Instituto de Epilepsia y Parkinson del eje cafetero
  • Walter Serna Serna Instituto de Epilepsia y Parkinson del eje cafetero
  • Genaro Daza Santacoloma Instituto de Epilepsia y Parkinson del eje cafetero http://orcid.org/0000-0002-1429-5925

DOI:

https://doi.org/10.22517/23447214.12851

Keywords:

Información mutua, optimización, procesamiento de imágenes, registro de imágenes

Abstract

the image registration in medical application have become very important topic in recent decades, because the planning, intervention and subsequent analysis of these medical procedures are improved, such as of imaging guided surgery where is required an especial precision and reliability. Image registration consists in align the images so that they are oriented in the same way, so you can combine information from both images in the same space and the same visual representation The main objective of this work is to compare systematically some methodologies for multimodal image registration using measures based in mutual information (MI). We found that the image registration based MI is able to register multimodal and monomodal images with little susceptibility to changes in brightness or contrast in the image, but with high sensitivity to noise. We also established that optimization techniques based on particle swarms have better performance for the registration compared to other techniques analyzed. Furthermore, the technique of cubic interpolation (spatial transformation) improve the reliability of the algorithm in multimodal images.

Downloads

Download data is not yet available.

Downloads

Published

2016-12-30

How to Cite

Ospina Torres, G. A., Serna Serna, W., & Daza Santacoloma, G. (2016). Systematic comparison of mutual information-based methodologies for the multimodal register of medical images. Scientia Et Technica, 21(4), 342–351. https://doi.org/10.22517/23447214.12851

Issue

Section

Ciencias Básicas