Segmenting Brain MRI using Adaptive Mean Shift

Jimenez-Alaniz, J.R.; Pohl-Alfaro, Mauricio; Médina-Bañuelos, Verónica; Yáñez-Suárez, Oscar

doi:10.1109/iembs.2006.4398106

Cited by 4 publications

(4 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In recent years, the active-models-based segmentation methods have been widely employed in MRI segmentation and have achieved considerable success. 1988, M. Kass et al utilized parametric active contour methods for medical image segmentation, and then many contour-based and shape-based methods extensively studied and widely employed in medical image segmentation [13][14][15][16], [49][50]. J.…”

Section: Class Ii: Segmentation Methods Based On Statistical Theorymentioning

confidence: 99%

“…1988, M. Kass et al utilized parametric active contour methods for medical image segmentation. Since then, many active-models-based methods, containing contour-based or shape-based, have been extensively studied and widely employed in medical image segmentation [13][14][15][16]. Many segmentation methods, which aim at the unique characteristic of MRI (such as lesion segmentation [17][18], and volumetric study on MRI [19][20][21][22], have also been used in important applications in clinical studies.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A review of segmentation method for MR image

Gao

2010

2010 International Conference on Image Analysis and Signal Processing

View full text Add to dashboard Cite

Section: Class Ii: Segmentation Methods Based On Statistical Theorymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A review of segmentation method for MR image

Gao

2010

2010 International Conference on Image Analysis and Signal Processing

View full text Add to dashboard Cite

“…, where π(x) is the pilot density estimate obtained by first running mean shift with analysis bandwidth, σ o . They have found more use in smoothing type applications as reported in [25,19]. Variants have also been used in tracking scenarios, where the bandwidths are adapted in a task specific fashion (see [18,9], for example).…”

Section: Motivation and Backgroundmentioning

confidence: 99%

Anisotropic Agglomerative Adaptive Mean-Shift

Sawhney

Christensen

2014

Proceedings of the British Machine Vision Conference 2014

View full text Add to dashboard Cite

Mean Shift today, is widely used for mode detection and clustering. The technique though, is challenged in practice due to assumptions of isotropicity and homoscedasticity. Isotropic/scalar bandwidths tend to smooth anisotropic patterns and affect partition boundaries, while homoscedastic / global bandwidths are inappropriate when clusters (or modes) at different scales need to be identified.We present an adaptive Mean Shift methodology that allows for anisotropic clustering, through unsupervised local bandwidth selection. The bandwidth matrices evolve naturally, adapting locally through agglomeration, and in turn guiding further agglomeration. The online methodology is practical for low-dimensional feature spaces, preserving better detail and clustering salience. Additionally, conventional Mean Shift either critically depends on a per instance choice of bandwidth, or relies on offline methods which are inflexible and/or again data instance specific. The presented approach, due to its adaptive design, also alleviates this issuewith a default form performing generally well. The methodology though, allows for effective tuning of results.In the proposed approach, clusters arise on the fly, as a consequence of agglomeration of extant clusters. Local bandwidths which evolve anisotropically every iteration, are associated with each cluster; by design, all members of a cluster converge to the same local mode. By evolving as function of a cluster's aggregated trajectory points, these bandwidths are able to adapt to the underlying mode structure (shape, scale, orientation) -and in turn, guide future cluster trajectory and agglomeration. This results in robust mode detection and with increased partition saliency (Figs. 1, 2(a)). The supplementary presents a convergence proof when anisotropic bandwidths vary between Mean shift iterations, as is the case here. The approach involves running Mean Shift fixed point iterations at cluster levels, over a single data point per cluster. Starting out as trivial clusters (solitary data points), the clusters agglomerate between iterations. By algorithm design, clusters are merged only when they are tending towards the same mode. All member points of a cluster, u, which will eventually converge to a common local mode, share a common bandwidth, Σ u -referred to as the local bandwidth. This bandwidth evolves every iteration, adapting to the structure of the local mode and to an extent, its basin. The standard MS fixed point iteration, is reformulated through local bandwidth based decomposition, as a fixed point update over clusters :For ascertaining cluster merges, the data points in the vicinity of a cluster u's trajectory, u τ , are considered. If a data point, y, in vicinity of u τ , is ascertained to be heading to the same mode as u τ , then by transitivityall the members of its parent cluster, Π(y), are heading to that mode toothe clusters u and Π(y), can then be merged. The cluster which is higher up the mode (higher density) assimilates the other cluster into itself, thus accelera...

show abstract

“…Actually, this nonparametric method can automatically identify any number of arbitrarily shaped clusters. Moreover, this method was recently successfully used for the segmentation of brain MRI, which is a closely related problem [13]. We set the value of the smoothing parameter H to 2.…”

Section: Implementation Of the Algorithmmentioning

confidence: 99%

FuRIA: A Novel Feature Extraction Algorithm for Brain-Computer Interfaces using Inverse Models and Fuzzy Regions of Interest

Lotte

Lécuyer

Arnaldi

2007

2007 3rd International IEEE/EMBS Conference on Neural Engineering

View full text Add to dashboard Cite

In this paper, we propose a new feature extraction algorithm for Brain-Computer Interfaces (BCIs). This algorithm is based on inverse models and uses the novel concept of fuzzy Region Of Interest (ROI). It can automatically identify the relevant ROIs and their reactive frequency bands. The activity in these ROIs can be used as features for any classifier. A first evaluation of the algorithm, using a Support Vector Machine (SVM) as classifier, is reported on data set IV from BCI competition 2003. Results are promising as we reached an accuracy on the test set ranging from 85% to 86% whereas the winner of the competition on this data set reached 84%.

show abstract

Segmenting Brain MRI using Adaptive Mean Shift

Cited by 4 publications

References 4 publications

A review of segmentation method for MR image

A review of segmentation method for MR image

Anisotropic Agglomerative Adaptive Mean-Shift

FuRIA: A Novel Feature Extraction Algorithm for Brain-Computer Interfaces using Inverse Models and Fuzzy Regions of Interest

Contact Info

Product

Resources

About