Brain segmentation

Patient specific brain segmentation with applications

Patient specific brain segmentation refers to the automatic labelling of the different tissues/structures in the brain—e.g. white matter, grey-matter, cerebrospinal fluid using information obtained from one or more imaging modalities; e.g. different MRI techniques and CT. The resulting segmentation can then be used to construct patient-specific electromagnetic and biomechanical models. The former find use in applications such as localizing the source of epileptic seizures or localizing hyperthermia treatment for tumours. The latter find use in applications such as the modelling of tissue deformation during and after surgical intervention.

The aim of this project is to develop state-of-the-art algorithms for automatically and accurately performing patient specific brain segmentation. The efficacy of each algorithm will be determined using real clinical data. Automatic segmentation will be quantitatively compared to manual segmentations performed by several radiologists. Efficacy will also be assessed indirectly in terms of the measurable impact on performance in several applications; e.g. localizing the source of epileptic seizures from EEG measurements.

Our Approach

The majority of methods that have been proposed for automated segmentation of brain tissues are based on statistical parametric models. The MPM-MAP (Maximizer of the posterior marginals- Maximum a posteriori) algorithm [1] exemplifies this approach. It implements Bayesian segmentation based on non-rigid registration of the atlas[AM1] . The algorithm uses Expectation-Maximization (EM) for estimation of model parameters and Hidden Markov Random Fields (HMRF) for spatial coherences. Two other examples are KVL (K. Van Leemput) [2] and CGMM (Constrained Gaussian mixture model) [3] which use maximum a posteriori probability (MAP) or maximum likelihood (ML) method for the estimation of model parameters. A drawback with these approaches is that it is difficult to integrate pixel spatial information with multi-valued pixel information (e.g. when several different MR scans have been acquired). This is because the HMRF is itself hard to implement[AM2]  in high dimensional feature space.

Mean-shift (MS) segmentation overcomes this drawback. Mean-shift [4, 5] is a non-parametric technique used to estimate the modes of the multivariate distribution underlying a feature space. It does not require any prior information to initialize the position of the clusters and also does not constraint the shape of the clusters. Mean-shift segmentation involves concatenating both the spatial and range domains of an image and identifying modes in this multidimensional joint spatial-range feature space. The only free parameter is the kernel size (called the bandwidth). If the chosen value is too small then insignificant modes are detected (over-clustering). If it is too large then significant modes can be merged (under-clustering). Several methods [6] are available for the estimation of a single fixed bandwidth. However, the use of a single fixed bandwidth has the drawback that it can yield under- or over-clustering when the feature space has significantly different local characteristics across the space. Variable or adaptive bandwidth methods have been proposed [7] to overcome this drawback. Such methods involve determining the bandwidth value for each feature point by using the pilot density estimate[AM3] . In [8] it was shown that adaptive mean-shift clustering, in which the pilot density estimate is based on the nearest neighborhood, performs better than a single fixed bandwidth in high dimensional space. Nevertheless the bandwidth value defined in terms of the Euclidean distance to the  farthest feature point from the kernel center can be biased by outliers [9].

Current Developments

We present a novel adaptive mean shift (AMS) algorithm for the segmentation of tissues in magnetic resonance (MR) brain images. In particular we introduce a novel Bayesian approach for the estimation of the adaptive kernel bandwidth and investigate its impact on segmentation accuracy. The approach includes modeling the distributions of variances of k-nearest neighbor (kNN) features and fitting Gamma distribution functions to these. It is suitable when there is no knowledge of the underlying feature distribution. In our variation we used this idea locally for adaptive bandwidth estimation.We studied the three class problem where the brain tissues are segmented into white matter, gray matter and cerebrospinal fluid. The segmentation experiments were performed on both multi-modal simulated and real patient T1-weighted MR volumes with different noise characteristics and spatial inhomogeneities. The performance of the algorithm was evaluated relative to several competing methods using real and synthetic data. Our results demonstrate the efficacy of the proposed algorithm and that it can outperform widely used brain segmentation methods; kmeans and HMRF-EM especially when the noise and spatial intensity inhomogeneities are high.

Outlook

The resulting segmentation by our proposed method BAMS (Bayesian based adaptive Mean-Shift) permits not only the quantitative characterization of these tissues (e.g. in the study of Alzheimer’s disease and multiple sclerosis), but also the construction of patient specific models of tissue conductivity (e.g. for application in EEG source localization in epilepsy patients) or biomechanical properties (e.g. for application in surgical simulation).

Research team

MedTech West partner

Qaiser Mahmood, PhD Student, Department of Signal Processing and Biomedical Engineering, S2, Chalmers University of Technology.

Technical Research Partners

Dr. Andrew Mehnert, Department of Signal Processing and Biomedical Engineering, S2, Chalmers University of Technology.

Prof. Mikael Persson, Department of Signal Processing and Biomedical Engineering, S2, Chalmers University of Technology.

Dr. Artur Chodorowski, Department of Signal Processing and Biomedical Engineering, S2, Chalmers University of Technology.

Dr. Fredrik Edelvik, Department of Computational Engineering and Design, FCC, Chalmers University of Technology.

Clinical partners

Dr. Johanna Gellermann, MD, Berlin.

Prof. Mikael Elam, Institute of Neuroscience and Physiology, Clinical Neurophysiology, Sahlgrenska University Hospital

Dr. Anders Hedström, MD, Institute of Neuroscience and Physiology, Clinical Neurophysiology, Sahlgrenska University Hospital

Ass. Prof. Magnus Thordstein, Institute of Neuroscience and Physiology, Clinical Neurophysiology, Sahlgrenska University Hospital

Contact

Henrik Mindedal, MedTech West, henrik.mindedal@medtechwest.se

 

References

[1] J. L. Marroquín, B. C. Vemuri, S. Botello, and F. Calderon, “An accurate and efficient Bayesian method for automatic segmentation of brain MRI,” IEEE Trans. Med. Imag., vol.21, no.8, Aug. 2002, pp. 934–944.

[2] K. Van Leemput, F. Maes, D. Vandeurmeulen, and P. Suetens, “Automated model-based tissue classification of MR images of the brain,” IEEE Trans. Med. Imag., vol.18, no.10 , 1999, pp. 897–908.

[3] H. Greenspan, A. Ruf and J. Goldberger, “Constrained Gaussian mixture model framework for automatic segmentation of MR brain images,” IEEE Trans. Med. Imag.,  vol. 25, no. 9, Sep. 2006, pp. 1233–1245.

[4] K. Fukunaga and L. Hostetler, “The estimation of the gradient of a density function, with applications in pattern recognition,” IEEE Trans Inf. Theory, vol.21, no.1, 1975, pp. 32-40.

[5] D. Comaniciu and P. Meer, “Mean Shift: A robust approach toward feature space analysis,” IEEE Trans. Pattern Analysis Machine Intell., vol. 24, no. 5, 2002, pp. 603-619.

[6] M. C. Jones, J.S Marron and S. J. Sheather, “A brief survey of bandwidth selection for density estimation,’’ Journal of the American Statistical Association, vol.91, no.433, pp.401-407.

[7] D. Comaniciu, V. Ramesh , P. Meer, “The variable bandwidth mean-shift and data-driven scale selection,” in ICCV,  2001, pp.438-445.

[8] B. Georgescu, I. Shimshoni, and P. Meer, “Mean-shift based clustering in high dimensions: A texture classification example,” in ICCV, 2003, pp. 456–463.

[9] A. G. Bors and N. Nasios, “Kernel bandwidth estimation for nonparametric modeling,” IEEE Trans.SMC, 2009, pp.1543 – 1555.

Publication

Qaiser Mahmood, Artur Chodorowski, Andrew Mehnert, Mikael Persson, “A Novel Bayesian Approach to Adaptive Mean Shift Segmentation of Brain Images,’’ IEEE CBMS, 2012.