Fully Automatic Brain Tumor Segmentation from Multiple MR Sequences using Hidden Markov Fields and Variational

GAIN is an enhanced version of the original Grouping Artificial Immune Network that was developed for fully automated MRI brain segmentation. The model captures the main concepts by which the immune system recognizes pathogens and models the process in a numerical form. GAIN was adapted to support a variable number of input patterns for training and segmentation of tumors in MRI brain images and adapted to train on multiple images. The model was demonstrated to operate with multi-spectral MR data with an increase in accuracy compared to the single spectrum case. Using the BRATS High Grade 2013 dataset with the 2012 tissue labels for Edema and Tumor, the model’s Dice scores were compared to published results and proved to be as accurate as the best methods. Using the 4 labels from the BRATS 2013 data sets, a Dice overlap of 73% for the complete tumor region and 64% for the enhancing tumor region were obtained for the high grade BRATS images when applying preand post-processing. This was attained with speed optimizations allowing segmentation at 21s per case with post-processing of all 4 tissues.

[1]  Zhuowen Tu,et al.  Auto-Context and Its Application to High-Level Vision Tasks and 3D Brain Image Segmentation , 2010, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[2]  F. Barkhof,et al.  Radiotherapy response of cerebral metastases quantified by serial MR imaging , 2005, Journal of Neuro-Oncology.

[3]  Gilles Celeux,et al.  EM procedures using mean field-like approximations for Markov model-based image segmentation , 2003, Pattern Recognit..

[4]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[5]  G. Clark,et al.  Reference , 2008 .

[6]  Michael J Ackerman,et al.  Engineering and algorithm design for an image processing Api: a technical report on ITK--the Insight Toolkit. , 2002, Studies in health technology and informatics.

[7]  Brian B. Avants,et al.  The optimal template effect in hippocampus studies of diseased populations , 2010, NeuroImage.

[8]  Arno Klein,et al.  A reproducible evaluation of ANTs similarity metric performance in brain image registration , 2011, NeuroImage.

[9]  Ben Glocker,et al.  Decision Forests for Tissue-Specific Segmentation of High-Grade Gliomas in Multi-channel MR , 2012, MICCAI.

[10]  Stephen M. Smith,et al.  SUSAN—A New Approach to Low Level Image Processing , 1997, International Journal of Computer Vision.

[11]  Brian B. Avants,et al.  N4ITK: Improved N3 Bias Correction , 2010, IEEE Transactions on Medical Imaging.

[12]  Robert J. Ogg,et al.  Efficacy of texture, shape, and intensity features for robust posterior-fossa tumor segmentation in MRI , 2009, Medical Imaging.

[13]  Ronald Marsh,et al.  Fractal analysis of tumor in brain MR images , 2003, Machine Vision and Applications.

[14]  Wei-Yin Loh,et al.  Classification and Regression Tree Methods , 2008 .

[15]  Stefan Bauer,et al.  Fully Automatic Segmentation of Brain Tumor Images Using Support Vector Machine Classification in Combination with Hierarchical Conditional Random Field Regularization , 2011, MICCAI.

[16]  Brian B. Avants,et al.  An Open Source Multivariate Framework for n-Tissue Segmentation with Evaluation on Public Data , 2011, Neuroinformatics.

[17]  B Scherrer,et al.  Fully Bayesian joint model for MR brain scan tissue and structure segmentation. , 2008, Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention.

[18]  Min Chen,et al.  Multi-parametric neuroimaging reproducibility: A 3-T resource study , 2011, NeuroImage.

[19]  Antonio Criminisi,et al.  Decision Forests for Computer Vision and Medical Image Analysis , 2013, Advances in Computer Vision and Pattern Recognition.

[20]  A. Ciarmiello,et al.  Automated segmentation and measurement of global white matter lesion volume in patients with multiple sclerosis. , 2000, Journal of magnetic resonance imaging : JMRI.

[21]  R. Kikinis,et al.  Automated segmentation of MR images of brain tumors. , 2001, Radiology.

[22]  Dong Hye Ye,et al.  Context-sensitive Classication Forests for Segmentation of Brain Tumor Tissues , 2012 .

[23]  Antonio Criminisi,et al.  Decision Forests: A Unified Framework for Classification, Regression, Density Estimation, Manifold Learning and Semi-Supervised Learning , 2012, Found. Trends Comput. Graph. Vis..

[24]  Stephen Gould,et al.  Multiclass pixel labeling with non-local matching constraints , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[25]  Nelly Gordillo,et al.  State of the art survey on MRI brain tumor segmentation. , 2013, Magnetic resonance imaging.

[26]  Nikos Komodakis,et al.  Performance vs computational efficiency for optimizing single and dynamic MRFs: Setting the state of the art with primal-dual strategies , 2008, Comput. Vis. Image Underst..

[27]  Khan M. Iftekharuddin,et al.  Efficacy of Texture, Shape, and Intensity Feature Fusion for Posterior-Fossa Tumor Segmentation in MRI , 2011, IEEE Transactions on Information Technology in Biomedicine.

[28]  F Barkhof,et al.  Interobserver variability in the radiological assessment of response to chemotherapy in glioma , 2003, Neurology.

[29]  Michael I. Jordan,et al.  On Discriminative vs. Generative Classifiers: A comparison of logistic regression and naive Bayes , 2001, NIPS.

[30]  Jitendra Malik,et al.  Representing and Recognizing the Visual Appearance of Materials using Three-dimensional Textons , 2001, International Journal of Computer Vision.

[31]  R P Velthuizen,et al.  MRI segmentation: methods and applications. , 1995, Magnetic resonance imaging.

[32]  D. Mumford,et al.  Optimal approximations by piecewise smooth functions and associated variational problems , 1989 .

[33]  Kenneth I. Laws,et al.  Rapid Texture Identification , 1980, Optics & Photonics.

[34]  G. Dai,et al.  Novel membrane‐permeable contrast agent for brain tumor detection by MRI , 2010, Magnetic resonance in medicine.

[35]  Cornelis H. Slump,et al.  MRI modalitiy transformation in demon registration , 2009, 2009 IEEE International Symposium on Biomedical Imaging: From Nano to Macro.

[36]  Pascal Fua,et al.  SLIC Superpixels Compared to State-of-the-Art Superpixel Methods , 2012, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[37]  L. Clarke,et al.  Monitoring brain tumor response to therapy using MRI segmentation. , 1997, Magnetic resonance imaging.

[38]  S. Bauer,et al.  A survey of MRI-based medical image analysis for brain tumor studies , 2013, Physics in medicine and biology.

[39]  R. Kimmel,et al.  Geodesic Active Contours , 1995, Proceedings of IEEE International Conference on Computer Vision.

[40]  D. Rubin,et al.  Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .

[41]  Ben Glocker,et al.  Classification Forests for Semantic Segmentation of Brain Lesions in Multi-channel MRI , 2013 .

[42]  J. Sethian,et al.  FRONTS PROPAGATING WITH CURVATURE DEPENDENT SPEED: ALGORITHMS BASED ON HAMILTON-JACOB1 FORMULATIONS , 2003 .

[43]  J. Gee,et al.  N4ITK: Nick's N3 ITK Implementation For MRI Bias Field Correction , 2010, The Insight Journal.

[44]  C. Meltzer,et al.  Brain tumor volume measurement: comparison of manual and semiautomated methods. , 1999, Radiology.

[45]  Jayaram K. Udupa,et al.  New variants of a method of MRI scale standardization , 2000, IEEE Transactions on Medical Imaging.

[46]  Colin Studholme,et al.  A Supervised Patch-Based Approach for Human Brain Labeling , 2011, IEEE Transactions on Medical Imaging.

[47]  Alan L. Yuille,et al.  Efficient Multilevel Brain Tumor Segmentation With Integrated Bayesian Model Classification , 2008, IEEE Transactions on Medical Imaging.