Download presentation
Presentation is loading. Please wait.
1
EMAtlasBrainClassifier By Kilian Maria Pohl pohl@mit.edu http://www.csail.mit.edu/~pohl
2
Kilian M. Pohl - 2 - Pipeline 2 Atlas Alignment 3 EM Segmentation 1 Intensity Normalization
3
Kilian M. Pohl - 3 - Intensity Normalization BeforeAfter
4
Kilian M. Pohl - 4 - Registration Non-Rigid Registration Based on Maxwell Demons [Guimond 99] Accurate but computationally expensive Affine Registration Based on Mutual Information [van Leemput 99] Robust but generally not as accurate Joint Registration & Segmentation Based on work by [Pohl 05] Robust, accurate, and computationally expensive
5
Kilian M. Pohl - 5 - Atlas Alignment Pre-selected Subject Training Subjects Segmentations Spatial Prior Registration Resampling
6
Kilian M. Pohl - 6 - Segmentation
7
Kilian M. Pohl - 7 - Requirements Minimum of 1 Gig of RAM (1.5 h on PC) Aligned T1 and T2 of the whole head Optimized for –isotropic T2 (Dimension 0.9375mm x 0.9375mm x 3mm ) –SPGR (Dimension 0.9375mm x 0.9375mm x 1.5mm ) For further information see http://....
8
Kilian M. Pohl - 8 - Segmentation of 31 Structures
9
Kilian M. Pohl - 9 - Define Model Labelmap T Inhomogeneity B Parameter Data Image I Registration R Shape S
10
Kilian M. Pohl - 10 - Define Model Inhomogeneity B Labelmap T Inhomogeneity B Parameter Data Image I Registration R Shape S Hierarchy H
11
Kilian M. Pohl - 11 - Observed Data (ROI) EM Hierarchical Implementation Image Prior HierarchyLabelmap
12
Kilian M. Pohl - 12 - Level 1 Prior Information IMAGE BGICC CSFGMWM
13
Kilian M. Pohl - 13 - Level 2 IMAGE ICC Current Parameter CSFGMWM ROI
14
Kilian M. Pohl - 14 - Segmentation of 31 Structures
15
Kilian M. Pohl - 15 - The EM Algorithm Expectation Step: Calculate lower bound Maximization Step: Find maxima of bound Graph from Minka
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.