A taxonomy of mutual information in medical image registration

Borvornvitchotikarn, Thuvanan; Kurutach, Werasak

doi:10.1109/iwssip.2016.7502712

Cited by 5 publications

(2 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The y-axis represents the number of parameters of a given model, while the x-axis represents the running time model performance effectively. Some simple linear or filtering methods can also be used to filter features [3,11,15], such as point mutual information (PMI), PCA, Gaussian filtering, etc. These methods are mainly applied in the image pre-processing.…”

Section: Related Workmentioning

confidence: 99%

A stereo spatial decoupling network for medical image classification

You

Tian

et al. 2023

Complex Intell. Syst.

View full text Add to dashboard Cite

Deep convolutional neural network (CNN) has made great progress in medical image classification. However, it is difficult to establish effective spatial associations, and always extracts similar low-level features, resulting in redundancy of information. To solve these limitations, we propose a stereo spatial discoupling network (TSDNets), which can leverage the multi-dimensional spatial details of medical images. Then, we use an attention mechanism to progressively extract the most discriminative features from three directions: horizontal, vertical, and depth. Moreover, a cross feature screening strategy is used to divide the original feature maps into three levels: important, secondary and redundant. Specifically, we design a cross feature screening module (CFSM) and a semantic guided decoupling module (SGDM) to model multi-dimension spatial relationships, thereby enhancing the feature representation capabilities. The extensive experiments conducted on multiple open source baseline datasets demonstrate that our TSDNets outperforms previous state-of-the-art models.

show abstract

Section: Related Workmentioning

confidence: 99%

A stereo spatial decoupling network for medical image classification

You

Tian

et al. 2023

Complex Intell. Syst.

View full text Add to dashboard Cite

show abstract

“…For the non-rigid image registration, the best implementation of MI is not trivial because it is a global measure, therefore its local estimation is difficult and using MI as the distance measure increasing the non-convexity of registration problem [6]. In [7], the authors provide a taxonomy for MI which summarises several variants of MI and their limitations.…”

Section: Introductionmentioning

confidence: 99%

Multi-modality image registration using the decomposition model

Ibrahim¹,

Chen²

2017

AIP Conference Proceedings

View full text Add to dashboard Cite

In medical image analysis, image registration is one of the crucial steps required to facilitate automatic segmentation, treatment planning and other application involving imaging machines. Image registration, also known as image matching, aims to align two or more images so that information obtained can be compared and combined. Different imaging modalities and their characteristics make the task more challenging. We propose a decomposition model combining parametric and non-parametric deformation for multi-modality image registration. Numerical results show that the normalised gradient field perform better than the mutual information with the decomposition model.

show abstract